Dissecting the Telomere-Independent Pathways
Underlying Human Cellular Senescence
By
Emilie Marie Isabelle Rovillain
A thesis submitted to the University College of London
for the degree of Doctor of Philosophy
Department of Neurodegenerative diseases
Institute of Neurology
UCL
Queen Square
London WC1 3BG
2010
ABSTRACT
Cellular senescence is an irreversible program of cell cycle arrest triggered in normal
somatic cells in response to a variety of intrinsic and extrinsic stimuli including telomere
attrition, DNA damage, physiological stress and oncogene activation.
Finding that inactivation of the pRB and p53 pathways by SV40-LT antigen cooperates
with hTERT to immortalize cells has allowed us to use a thermolabile mutant of SV40LT to develop human fibroblasts where the cells are immortal if grown at 34 oC but
undergo an irreversible growth arrest within 5 days at 38oC. When these cells cease
dividing, senescence-associated-β-galactosidase (SA-β-Gal) activity is induced and the
growth-arrested cells have many features of senescent cells.
Since these cells growth-arrest in a synchronous manner, I have used Affymetrix
expression profiling to identify the genes differentially expressed upon senescence. This
identified 816 up- and 961 down-regulated genes whose expression was reversed when
growth arrest was abrogated. I have shown that senescence was associated with activation
of the NF-B pathway and up-regulation of a number of senescence-associated-secretoryproteins including IL6. Perturbation of NF-κB signalling either by direct silencing of NFB subunits or by upstream modulation overcame growth-arrest indicating that activation
of NF-B signalling has a causal role in promoting senescence.
I also applied a retroviral shRNA screen covering ~10,000 genes to the same cell model.
Overlapping with the microarray data revealed particularly interesting targets, such as
LTBP3 and Layilin. Finally, I profiled micro-rna expression. 15 of the top micro-rnas
down-regulated upon senescence were chosen to express in the HMF3A system. 6 of
them were able to bypass the growth-arrest.
In conclusion, my work has uncovered novel markers involved in senescence as well as
identifying that both activation of p53 and pRb pathway result in activation of NF-B
signalling which promotes senescence. Both results lead to a better understanding of
senescence and its pathways.
2
TABLE OF CONTENTS
ABSTRACT .................................................................................................................... 2
TABLE OF CONTENTS ................................................................................................ 3
LIST OF FIGURES ....................................................................................................... 14
LIST OF TABLES ........................................................................................................ 17
LIST OF ABBREVIATIONS ........................................................................................ 18
ENCLOSED UNBOUND MATERIAL ......................................................................... 22
STATEMENT CONCERNING COLLABORATIONS ................................................. 22
ACKNOWLEDGEMENTS ........................................................................................... 22
1
INTRODUCTION ................................................................................................. 24
1.1
REPLICATIVE SENESCENCE DISCOVERY ............................................. 24
1.2
DEFINITION OF SENESCENCE ................................................................. 25
1.3
TELOMERE INDUCED SENESCENCE ...................................................... 25
1.3.1 Telomeres .................................................................................................. 25
1.3.2 Telomeres and DDR ................................................................................... 27
1.3.3 hTERT ....................................................................................................... 29
1.3.4 Telomerase and tumourigenesis.................................................................. 30
1.4
DNA DAMAGE INITIATED SENESCENCE ............................................... 30
1.5
ONCOGENE-INDUCED SENESCENCE ..................................................... 32
1.6
CANCER AND SENESCENCE .................................................................... 33
1.7
AGEING AND SENESCENCE ..................................................................... 33
1.8
PATHWAYS OF SENESCENCE .................................................................. 34
1.8.1 The p53 pathway ........................................................................................ 34
1.8.1.1
1.8.1.2
1.8.1.3
1.8.1.4
1.8.1.5
1.8.1.6
1.8.1.7
TP53 gene and p53 protein ................................................................. 35
Functions of the p53 protein ............................................................... 35
Regulation of p53 activity................................................................... 36
p19Arf protein...................................................................................... 38
Oncogenic Ras ................................................................................... 40
p21WAF1/Cip1/Sdi1 ................................................................................... 40
p53 family: p63 and p73 proteins ........................................................ 41
1.8.2 The pRb pathway ....................................................................................... 42
1.8.2.1
Cell cycle, cyclins and CDKs ............................................................. 42
1.8.2.2
1.8.2.3
1.8.2.4
1.8.2.5
1.8.2.6
1.8.2.7
1.8.2.8
CDK inhibitors ................................................................................... 42
CDKs and E2F ................................................................................... 43
Rb family of proteins .......................................................................... 43
pRb gene ............................................................................................ 44
pRb discovery .................................................................................... 44
pRb function....................................................................................... 45
E2F .................................................................................................... 47
1.8.3 Common pathways ..................................................................................... 48
1.8.3.1
1.8.3.2
1.8.3.3
1.9
INK4A Locus ..................................................................................... 48
p16INK4a .............................................................................................. 49
p14ARF ................................................................................................ 50
DNA TUMOUR VIRUSES ........................................................................... 52
1.9.1 SV40 .......................................................................................................... 52
1.9.2 LT .............................................................................................................. 53
1.9.3 Adenovirus Type 5 ..................................................................................... 55
1.9.3.1
1.9.3.2
E1A .................................................................................................... 55
E1B .................................................................................................... 56
1.9.4 HPV Type 16 ............................................................................................. 56
1.9.4.1
1.9.4.2
1.10
E7 ...................................................................................................... 57
E6 ...................................................................................................... 58
SASP: SENESCENCE-ASSOCIATED SECRETORY PHENOTYPE AND
ROS: REACTIVE OXYGEN SPECIES .................................................................... 58
1.11
NF-B PATHWAY ....................................................................................... 60
1.11.1 Introduction ............................................................................................... 60
1.11.2 NF-κB family ............................................................................................. 61
1.11.3 Activation .................................................................................................. 62
1.11.4 Inhibition ................................................................................................... 63
1.11.4.1
1.11.4.2
The IκB family ............................................................................... 63
IκB kinase: IKK............................................................................. 63
1.11.5 Canonical NF-B pathway ......................................................................... 65
1.11.6 Non-canonical pathway .............................................................................. 66
1.11.7 NF-B and cancer ...................................................................................... 68
1.11.8 NF-B, senescence and ageing ................................................................... 68
1.12
MICRO-RNAS .............................................................................................. 69
4
1.12.1 Introduction ............................................................................................... 69
1.12.2 MiRNA and siRNA .................................................................................... 70
1.12.3 Biogenesis .................................................................................................. 71
1.12.4 Mechanism of MiRNA regulation .............................................................. 71
1.12.5 Micro-RNAs and cancer ............................................................................. 73
1.12.6 Micro-RNA and senescence ....................................................................... 74
1.13
MODEL OF STUDY: HMF3A CELLS ......................................................... 75
1.13.1 Reconstitution of WT LT activity in the HMF3A system alone .................. 76
1.13.2 Refinement of the HMF3A system by introduction of the murine ecotropic
receptor ................................................................................................................. 78
1.14
ABROGATION OF THE P53 PATHWAY ................................................... 78
1.15
ABROGATION OF THE PRB PATHWAY .................................................. 78
1.15.1 Inactivation of the INK4A Locus ............................................................... 79
1.15.1.1
1.15.1.2
1.15.1.3
1.16
2
Knockdown of p14ARF by ShRNA .................................................. 79
Knockdown of p16INK4a by ShRNA ................................................ 79
Constitutive Expression of Bmi-1 ................................................... 80
AIM OF THE THESIS................................................................................... 80
MATERIAL AND METHODS ............................................................................. 83
2.1
MAMMALIAN CELL CULTURE ................................................................ 83
2.1.1 Cell lines and Culture ................................................................................. 83
2.1.2 Cell media .................................................................................................. 83
2.1.3 Cell Culture Conditions .............................................................................. 83
2.1.4 Sub-Culturing of Cells ............................................................................... 84
2.1.5 Preservation of Cells .................................................................................. 84
2.1.6 Recovery of Frozen Cells ........................................................................... 84
2.2
RETROVIRAL AND LENTIVIRAL INFECTIONS ..................................... 85
2.2.1 Retroviral and Lentiviral constructs ............................................................ 85
2.2.2 Viral Packaging and infections ................................................................... 86
2.2.2.1
2.2.2.2
2.2.2.3
2.3
Packaging of Retroviral Constructs ..................................................... 86
Packaging of Lentiviral Constructs ..................................................... 87
Infection with viral supernatant and selection ..................................... 87
DNA MANIPULATION................................................................................ 88
5
2.3.1 Plasmid DNA Preparation .......................................................................... 88
2.3.1.1
2.3.1.2
Small Scale Plasmid Preparation ........................................................ 88
Large Scale Plasmid Preparation ........................................................ 89
2.3.2 DNA Quantification ................................................................................... 90
2.3.3 DNA-Agarose Gel Electrophoresis ............................................................. 90
2.3.4 DNA Sequencing ....................................................................................... 90
2.3.5 Cloning of PCR Products ........................................................................... 90
2.4
RNA MANIPULATION ................................................................................ 91
2.4.1 RNA Isolation ............................................................................................ 91
2.4.2 RNA Quantification ................................................................................... 91
2.5
PROTEIN ANALYSIS .................................................................................. 92
2.5.1 Preparation of Total Protein Extracts .......................................................... 92
2.5.2 Determination of Protein Concentration ..................................................... 92
2.5.3 Sodium-Dodecyl-Sulphate-Polyacrylamide-Gel-Electrophoresis ................ 93
2.5.4 Western Blotting of SDS-PAGE................................................................. 93
2.5.5 Antibodies Used ......................................................................................... 94
2.6
GROWTH CURVES ..................................................................................... 95
2.6.1 Cell line ..................................................................................................... 95
2.6.2 Protocol ..................................................................................................... 95
2.7
IRREVERSIBILITY ASSAYS ...................................................................... 95
2.7.1 Cell line ..................................................................................................... 95
2.7.2 Protocol ..................................................................................................... 96
2.8
GROWTH COMPLEMENTATION ASSAYS .............................................. 96
2.8.1 Cell line ..................................................................................................... 96
2.8.2 Complementation experiment ..................................................................... 96
2.9
SENESCENCE SPECIFIC EXPRESSION PROFILING ............................... 97
2.9.1 Cell line ..................................................................................................... 97
2.9.2 RNA preparation ........................................................................................ 97
2.9.3 RNA expression profiling........................................................................... 98
2.10
SENESCENCE SPECIFIC MIRNA EXPRESSION PROFILING ................. 98
6
2.10.1 Cell line ..................................................................................................... 98
2.10.2 Tissue culture ............................................................................................. 98
2.10.3 RNA preparation ........................................................................................ 99
2.10.4 Quality Control of RNA Samples and shipping to LC sciences ................. 100
2.10.5 Microarray analysis .................................................................................. 100
2.10.5.1
2.10.5.2
Pairwise Comparisons for micro-rna Microarray Analysis ............ 100
Expression analysis and normalization .......................................... 101
2.10.6 Individual miRNA validation in vitro ....................................................... 101
2.10.6.1
2.10.6.2
2.10.6.3
2.11
Candidates miR-Vec Clones ......................................................... 102
Preparation of clones and sequence checking ................................ 102
HMF3A Growth Complementation Assay in plates ...................... 102
SHRNA SCREENING ................................................................................. 103
2.11.1 Cell line ................................................................................................... 103
2.11.2 RNAi library ............................................................................................ 103
2.11.3 Virus packaging ....................................................................................... 103
2.11.4 Sensitivity of the model............................................................................ 104
2.11.5 Titration of Phoenix Eco viral Supernatants.............................................. 104
2.11.6 Experiment planning for each plasmids pools ........................................... 105
2.11.7 Confidence intervals................................................................................. 105
2.11.8 Genomic DNA extraction ......................................................................... 106
2.11.9 TOPO cloning and sequencing ................................................................. 107
2.12
3
PRIMARY CELLS IMMORTALIZATION ................................................. 108
CELL MODEL AND GROWTH COMPLEMENTATION ................................. 109
3.1
CREATION OF CLONAL CELL LINES DERIVED FROM HMF3A CELLS
AND GROWTH COMPLEMENTATION ASSAYS............................................... 109
3.1.1 Objectives ................................................................................................ 109
3.1.2 Refinement of the HMF3A cells by clonal selection ................................. 109
3.1.2.1
3.1.2.2
3.1.2.3
Growth curves .................................................................................. 110
Complementation with E7 and E1A .................................................. 112
Irreversibility.................................................................................... 115
3.1.3 Reconstitution of WT LT antigen in HMF3AEcoR and CL3EcoR cells ......... 122
3.1.4 Abrogation of the p53 pathway in HMF3AEcoR and CL3EcoR cells ............. 126
3.1.5 Abrogation of the pRb pathway in HMF3AEcoR and CL3EcoR cells ............ 128
7
3.1.5.1
3.1.5.2
3.2
Constitutive Expression of Ad 5 E1A and HPV16 E7 ....................... 131
Constitutive ectopic expression of E2F-DB mutant ........................... 136
DISCUSSION.............................................................................................. 138
3.2.1 Cellular senescence is p53-dependant process in the HMF3A cells ........... 139
3.2.2 Senescence is a p21CIP1/WAF1/Sdi1-Dependent Process in the HMF3A Cells 141
3.2.3 Inactivation of the pRb pathway in the HMF3A cells ............................... 142
3.2.3.1 p16INK4a inactivation in the HMF3A Cells ........................................ 142
3.2.3.2 Bmi-1 Activity in the HMF3A Cells ................................................. 143
3.2.3.3 Ectopic Expression of E1A and E7 ................................................... 143
3.2.3.4 Possible Mechanisms by which E1A and E7 Bypass the Conditional
HMF3A Growth Defect ................................................................................... 144
3.2.3.5 p14ARF is not necessary between the p16-pRb and p21-p53 Pathways .....
......................................................................................................... 145
3.2.3.6 E2F-DB bypass the conditional growth arrest by repressing the pRb
pathway ......................................................................................................... 146
3.2.4 p16-pRb does not always act downstream of p53-p21 to induce senescence ...
................................................................................................................ 147
4
ACTIVATION OF THE NF-κB SIGNALLING PROMOTES CELLULAR
SENESCENCE ........................................................................................................... 148
4.1
SENESCENCE SPECIFIC GENE EXPRESSION RESULTS ..................... 148
4.1.1 Objectives ................................................................................................ 148
4.1.2 Why use Microarray Analysis?................................................................. 149
4.1.3 Which Microarray Technology? ............................................................... 151
4.1.4 Microarray Strategy ................................................................................. 152
4.1.5 Microarray procedure ............................................................................... 156
4.1.6 Microarray results .................................................................................... 156
4.1.7 Validation of Microarray Data.................................................................. 162
4.1.7.1
4.1.7.2
Why Validate?.................................................................................. 162
Real time validation of expression data using the BioTrove Open Arrays
......................................................................................................... 162
4.1.8 Comparison of genes differentially expressed upon senescence with the
meta-signature of genes over-expressed in cancer ................................................ 165
4.2
BIOLOGICAL VALIDATION BY LENTIVIRAL SILENCING OR ECTOPIC
EXPRESSION......................................................................................................... 167
4.2.1 Objectives ................................................................................................ 167
8
4.2.2 Up-regulated genes upon senescence: Does Silencing bypass the growth
arrest? 167
4.2.2.1
4.2.2.2
4.2.2.3
4.2.2.4
4.2.2.5
4.2.2.6
4.2.2.7
4.2.2.8
4.2.2.9
CLCA2 silencing bypassed senescence at a low level ....................... 168
AK3L1 silencing bypassed senescence ............................................. 169
TRIB2 silencing bypassed senescence .............................................. 169
CDKN2A silencing bypassed senescence ......................................... 171
DAPK1 silencing was not sufficient to bypass senescence ................ 172
BLCAP silencing bypassed senescence at a low level ....................... 172
RUNX1 bypassed senescence ........................................................... 173
GRAMD3 silencing bypassed senescence ......................................... 173
SCN2A silencing was not sufficient to bypass senescence ................ 174
4.2.3 Down-regulated genes upon senescence: Does ectopic expression bypass the
growth arrest? ...................................................................................................... 174
4.2.3.1
4.2.3.2
4.2.3.3
4.2.3.4
4.2.3.5
4.2.3.6
4.2.3.7
4.2.3.8
4.3
HMGB2 ........................................................................................... 176
DEPDC1 .......................................................................................... 176
BUB1B ............................................................................................ 177
NEK2 ............................................................................................... 179
MELK .............................................................................................. 180
MLF1-IP two splice forms 88 and 401 .............................................. 181
DBF4, CDKN2C (p18) and PLK4 .................................................... 182
FOXM1 ............................................................................................ 182
NF-κB PATHWAY ACTIVATION UPON SENESCENCE IS CAUSAL TO
SENESCENCE ....................................................................................................... 189
4.3.1 Objectives ................................................................................................ 189
4.3.2 NF-κB pathway is activated upon senescence at the mRNA level ............. 189
4.3.2.1
4.3.2.2
Transcription factor motif matrix module ......................................... 191
NF-κB targets gene expression modulation ....................................... 191
4.3.3 Is the NF-κB pathway also activated at a protein level? ............................ 195
4.3.4 Is phosphorylation of RelA/p65 also induced? .......................................... 195
4.3.5 What happens if the NF-κB complex is inactivated? ................................. 197
4.3.5.1 RNAi mediated silencing of NF-κB subunits abrogates senescence
growth arrest .................................................................................................... 197
4.3.6 Modulation of the NF-κB pathway overcomes senescence growth arrest .. 200
4.3.6.1
4.3.6.2
4.3.6.3
4.3.6.4
TMEM9B and BCL2L1 silencing bypass senescence ....................... 200
Silencing of cEBPβ, BTG2 and TXNIP silencing bypass senescence 202
Ectopic expression of IKB-SR bypasses senescence ......................... 202
Ectopic expression of SIRT1 bypasses senescence............................ 207
4.3.7 NF-κB Activation is Causal to Senescence ............................................... 207
9
4.4
DISCUSSION.............................................................................................. 209
4.4.1 SASP: Senescence-Associated Secretory Phenotype ................................ 209
4.4.2 Senescence Down-Regulated Genes ......................................................... 210
4.4.3 Ectopic expression of down regulated genes rescues the growth arrest ..... 211
4.4.4 NF-κB Pathway Activation upon Senescence is causal to Senescence ...... 214
4.4.4.1
4.4.4.2
4.4.4.3
profile
4.4.4.4
5
AN
RNA
Senescence Up-Regulated Genes ...................................................... 215
Silencing of over-expressed genes bypasses the growth arrest .......... 216
Senescence expression profile reveals links with Cancer expression
217
NF-κB pathway is activated upon senescence ................................... 218
INTERFERENCE
SCREEN
IDENTIFIES
DOWNSTREAM
EFFECTORS OF THE P53-P21 AND P16-PRB PATHWAYS ................................... 222
5.1
RNAI INTERFERENCE SCREEN .............................................................. 222
5.1.1 Objectives ................................................................................................ 222
5.1.2 The Open Access RNAi project at UCL ................................................... 223
5.1.3 Which viral delivery system? Which library? ........................................... 223
5.1.4 ShRNA Screening Strategy ...................................................................... 228
5.1.5 Sensitivity of the model............................................................................ 228
5.1.6 Confidence intervals................................................................................. 230
5.1.7 Titration of Phoenix Eco viral Supernatants.............................................. 232
5.1.8 Primary screen in the HMF3A: Procedure ................................................ 232
5.1.9 ShRNA constructs sequence recovery ...................................................... 239
5.1.10 Results of the primary screen ................................................................... 240
5.2
IN VITRO VALIDATION OF THE SCREENING ....................................... 247
5.2.1 Overlap of the candidates of the shRNA screen with microarray data for
genes up-regulated upon senescence in CL3EcoR cells ........................................... 247
5.2.2 Optimisation of the GIPZ lentiviral library ............................................... 247
5.2.3 Secondary screen using lentiviral shRNA silencing .................................. 248
5.3
DISCUSSION.............................................................................................. 256
5.3.1 Sensitivity, Stringency and Saturation ...................................................... 256
5.3.2 Positive hits of the primary HMF3A retroviral shRNA screen .................. 258
5.3.3 Overlap with the microarray up-regulated genes reveals new targets ........ 259
10
5.3.4 In vitro validation of ATXN10, LAYN, LTBP3, SGTB and TMEM9B
silencing .............................................................................................................. 259
5.3.4.1
5.3.4.2
5.3.4.3
5.3.4.4
5.3.4.5
5.3.4.6
6
TMEM9B ......................................................................................... 259
LTBP3 ............................................................................................. 261
ATXN10 .......................................................................................... 264
LAYN .............................................................................................. 265
SGBT ............................................................................................... 265
TAOK1, RAS4A and ARMCX2....................................................... 267
ROLE OF MICRO-RNAS IN CELLULAR SENESCENCE................................ 270
6.1
SENESCENCE SPECIFIC MICRO-RNA DIFFERENTIAL EXPRESSION270
6.1.1 Objectives ................................................................................................ 270
6.1.2 Background to micro-RNA Expression Profiling Technology .................. 270
6.1.3 HMF3AEcoR: miRNA expression profiling experimental design ............. 271
6.1.4 Quality Control of RNA Samples ............................................................. 273
6.1.5 miRNAs senescence specific differential expression ................................ 273
6.1.6 Up-regulated micro-RNAs ....................................................................... 283
6.1.7 Down-regulated micro-RNAs ................................................................... 283
6.2
BIOLOGICAL VALIDATION BY GROWTH COMPLEMENTATION
ASSAY IN THE HMF3A CELLS ........................................................................... 283
6.2.1 Objectives ................................................................................................ 283
6.2.2 Validation by ectopic expression .............................................................. 284
6.2.2.1
6.2.2.2
miR Vec clones ................................................................................ 284
Sequencing of the MiRVec clones .................................................... 287
6.2.3 Complementation assay with the miR-Vec clones .................................... 287
6.2.3.1
6.2.3.2
6.2.3.3
6.2.3.4
MiR-18a, miR-130b, miR-372, miR-373 and Let7a .......................... 288
MiR-92b, miR-15a, miR-16, miR-195 and miR-25 ........................... 288
MiR-195, miR-218, miR-20b, miR-29b, miR-186 and miR-25 ....... 291
MiR-128, miR-423-5p and Let7g...................................................... 291
6.2.4 Overlapping with the microarray data and the shRNA screen ................... 294
6.2.4.1
6.2.4.2
6.2.4.3
6.2.4.4
6.2.4.5
6.2.4.6
MiR-25............................................................................................. 294
MiR-195 ........................................................................................... 297
MiR-218 ........................................................................................... 297
MiR-193b ......................................................................................... 297
MiR-186 ........................................................................................... 298
MiR-423-5p...................................................................................... 298
11
6.3
EXPRESSION PROFILING OF HMF3A CELLS IN WHICH GROWTH
ARREST WAS OVERCOME BY ECTOPIC EXPRESSION OF MIRS ................. 298
6.3.1 Objectives ................................................................................................ 298
6.3.2 Microarray Strategy ................................................................................. 299
6.3.3 Microarray procedure ............................................................................... 301
6.3.4 MiR-186 .................................................................................................. 301
6.3.5 MiR-195 .................................................................................................. 301
6.3.6 MiR-25 .................................................................................................... 302
6.3.7 MiR-218 .................................................................................................. 302
6.3.8 MiR-423-5p ............................................................................................. 303
6.3.9 MiR-372 .................................................................................................. 303
6.4
RAS INDUCED PREMATURE SENESCENCE ......................................... 303
6.4.1 Objectives ................................................................................................ 303
6.4.2 Strategy.................................................................................................... 304
6.4.3 Procedure ................................................................................................. 305
6.5
DISCUSSION.............................................................................................. 307
6.5.1 Up-regulated micro-RNAs ....................................................................... 307
6.5.1.1
6.5.1.2
MiR-34a ........................................................................................... 308
MiR-146a ......................................................................................... 309
6.5.2 Down-regulated micro-RNAs ................................................................... 310
6.5.2.1
6.5.2.2
6.5.2.3
6.5.2.4
6.5.2.5
6.5.2.6
MiR-25............................................................................................. 310
MiR-195 ........................................................................................... 312
MiR-218 ........................................................................................... 313
MiR-193b ......................................................................................... 314
MiR-186 ........................................................................................... 315
MiR-423-5p...................................................................................... 315
6.5.3 Expression profiling of HMF3A cells in which senescence has been
bypassed by ectopic expression of miRs .............................................................. 316
6.5.4 Expression of the miRs in 226L cells ....................................................... 317
6.5.5 RAS transformation of primary cells ........................................................ 318
6.5.6 Further work ............................................................................................ 322
7
SUMMARY AND FINAL DISCUSSION ........................................................... 324
7.1
SUMMARY OF RESULTS ......................................................................... 325
12
7.2
FUTURE DIRECTIONS.............................................................................. 326
7.2.1 Saturation of shRNA screen in CL3EcoR .................................................... 326
7.2.2 Secondary shRNA screen ......................................................................... 327
7.2.3 Ectopic expression validation by protein analysis ..................................... 327
7.2.4 FOXM1 ................................................................................................... 328
7.2.4.1
7.2.4.2
7.2.4.3
7.2.4.4
7.3
8
Which spliceform is important? ........................................................ 328
Which kinases regulate the activation of FOXM1? ........................... 329
What is the mechanism of action of FOXM1? .................................. 330
What causes the decreased expression of FOXM1 in cell senescence? ....
......................................................................................................... 330
FINAL REMARKS ..................................................................................... 331
REFERENCES .................................................................................................... 332
13
LIST OF FIGURES
Figure 1.1:
Replicative Senescence……..………………….……………..
27
Figure 1.2:
Telomere shortening …………………….……………………
29
Figure 1.3:
Causes and consequences of cellular senescence …..………...
32
Figure 1.4:
The p53 signalling pathway…………………………………...
38
Figure 1.5:
The pRb signaling pathway………..…………………………..
40
Figure 1.6:
NF-κB: the canonical pathway…………………………………
65
Figure 1.7:
NF-κB: the non canonical pathway ……………………...…….
68
Figure 1.8:
Micro-RNAs biogenesis………………………..….…………...
73
Figure 1.9:
Engineering of the CL3 EcoR cells…………………...…………..
78
Figure 3.1:
Clonal cell lines growth rates…………………………………..
112
Figure 3.2:
Growth complementation assay in the clonal cell lines………..
114
Figure 3.3:
Growth complementation assay in the clonal cell lines repeat…
115
Figure 3.4:
Figure 3.5:
EcoR
Irreversibility HMF3A
and CL3
EcoR
Irreversibility HMF3A
EcoR
cells: Photos…………
and CL3
EcoR
117
cells: Growth assays and
staining…………………………………………………………………………
120
Figure 3.6:
Induction of SA-β-galactosidase…………..……….…………...
122
Figure 3.7:
Complementation HMF3AEcoR
by ectopic
expression and
silencing...............................................................................................................
Figure 3.8:
EcoR
Complementation CL3A
RNAi
124
by ectopic expression and RNAi silencing
……………..……...…………………………………………………………….
125
Expression of LT in the HMF3AEcoR cells …………....………...
126
Figure 3.10: Expression of p53 in the HMF3AEcoR cells …...………………...
130
Figure 3.9:
Figure 3.11: Expression of p21CIP1/WAF1/Sdi1 in the HMF3AEcoR cells................. 131
Figure 3.12: Conserved regions of the DNA tumour Viruses………………....
EcoR
Figure 3.13: Expression of pRb in the HMF3A
133
cells................................... 135
Figure 3.14: Expression of E7 in the HMF3AEcoR cells...................................... 136
Figure 3.15: Expression of E2F-DB in the HMF3AEcoR cells............................. 138
Figure 4.1:
Cancer: a multistep process……..……........................................... 151
Figure 4.2:
HMF3AEcoR Microarray strategy……………………..…………... 154
14
Figure 4.3:
Microarray Strategy for the complementations….........................
156
Figure 4.4:
Validation of microarray data by real time qPCR……………….
165
Figure 4.5:
In vitro validation of up-regulated microarray targets by silencing
constructs ……………....………..……………………………………………… 171
Figure 4.6:
Silencing of GRAMD3 and SCN2A……………..…………….... 176
Figure 4.7:
In vitro validation of down-regulated microarray targets by ectopic
expression………………………..………………………………………………. 179
Figure 4.8:
FOXM1……..……………………………………………
Figure 4.9:
FOXM1 protein expression………………......................……….. 188
184-185
Figure 4.10: Ectopic expression of FOXM1 WT, FOXM1ΔNΔKEN and FOXM16K........................................................................................................................... 189
Figure 4.11: FOXM1
protein
expression
in
cells
expressing
FOXM1
WT,
FOXM1ΔNΔKEN and FOXM1-6K …..………………………………………... 191
Figure 4.12: Secretion of IL8 and IL6 by senescent cells……………..............
197
Figure 4.13: Increase in phosphorylation of RelA (Ser536) in senescent cells.. 199
Figure 4.14: Silencing of NF-κB transcription factor subunits………………... 200
Figure 4.15: Silencing of TMEM9B and BCL2L1…..………………………... 202
Figure 4.16: Silencing of cEBPβ…..………………………………………..… 204
Figure 4.17: Silencing of BTG2 and TXNIP…..……………………………… 205
Figure 4.18: Silencing of secreted proteins CCl26, IGFBP7, GDF15 and IL32 206
Figure 4.19: Ectopic expression of IKB-SR.……………………………..…… 207
Figure 4.20: Ectopic expression of SIRT1……………………………………...
209
Figure 5.1:
Mir-30 adapted shRNAmiR transcript design……...……………. 225
Figure 5.2:
pSM2 retroviral plasmid: design and features………………...…. 227
Figure 5.3:
pGIPZ lentiviral plasmid: design and features ……..……............ 228
Figure 5.4:
shRNA screen strategy……..……………………….……............ 230
Figure 5.5:
Screen sensitivity test……………........................………….…… 232
Figure 5.6:
Silencing of TMEM9B (mix)……..……………………………... 251
Figure 5.7:
Silencing of TMEM9B (individual)……..………………………. 252
Figure 5.8:
Silencing of LTBP3…………………………………………….... 253
Figure 5.9:
Silencing of ATXN10……………………………………………. 254
15
Figure 5.10: Silencing of SGTB and LAYN……………..………………….… 256
Figure 6.1:
miRs microarray profiling strategy ……...………………………. 273
Figure 6.2:
Differential miRs upon growth arrest and quiescence…...……..... 281
Figure 6.3:
Ectopic expression of miR-18a, miR-130b, miR-373 and miR-372 290
Figure 6.4:
Ectopic expression of miR-25, miR-92b, miR-195, miR-15a and miR-16a
……………………...……………………………………………. 291
Figure 6.5:
Ectopic expression of miR-29b, miR-20, miR-186, miR-193b and miR218……………………………………………………………….. 293
Figure 6.6:
Ectopic expression of miR-423-5p, miR-128, and Let7g………... 294
Figure 6.7:
Ectopic expression of miR-423-5p, miR-218 and miR-19...…….. 296
Figure 6.8:
Ectopic expression of miR-423-5p, miR-186, miR-20b, miR-193b, miR29b, miR25, miR-195 and miR-218…...….................................... 297
Figure 6.9:
Microarray profiling strategy of cells expressing miR-218, miR195, miR193b, miR-423-5p and miR-25 ………………………………….. 301
Figure 6.10: Edctopic expression of micro-RNAs in human breast epithelial cells 320
16
LIST OF TABLES
Table 4.1:
Senescence specific changes in gene expression…………
Table 4.2:
Senescence
specific
changes
in
gene
158-159
expression
with
complementation………………………………………….
161-162
Table 4.3:
Results of comparison Affymetrix with OpenArrayTM…...
167
Table 4.4:
Metasignatures
of
neoplastic
transformation
and
undifferentiated
cancer………………………………………………...........
167
Table 4.5:
Transcription factor motifs………………………………..
193
Table 4.6:
Senescence specific changes in NF-κB target genes expression with
complementation……………..………………………..………………….
194-195
Table 5.1:
shRNA screen confidence………………..……………….
232
Table 5.2:
Virus pools titration and supernatant volume used……….
234
Table 5.3:
Reseeding densities and number of growing colonies obtained after growth
complementation assay …………………….……..……...
236-239
Table 5.4:
Results of the screen ……………………………………..
242-246
Table 5.5:
Senescence specific changes with complementation for ATXN10, LAYN,
LTBP3, SGTB and TMEM9B …………………………..
261
Table 6.1:
Dual hybridization Analysis………………..……………
275
Table 6.2:
Raw microarray results upon growth arrest………...……
277-278
Table 6.3:
Raw microarray results upon quiescence………………..
279-280
Table 6.4:
Up-regulated micro-RNAs upon senescence ……….…..
282
Table 6.5:
Down-regulated micro-RNAs upon senescence ………..
283
Table 6.6:
Layout of of the primary BJ cells immortalization experiment
307
17
LIST OF ABBREVIATIONS
2D-DIGE
3D
6-FAM
aa
AEBSF
Ago
ALT
ampR
APC
APS
ARF
ATP
ATCC
Β-Gal
bHLH
BLAST
blastR
bp
BrdU
BSA
CaCl2
CDK
CDKI
cDNA
Cfu
ChIP
CIP
CR
C-terminal
DAPI
dd
DDR
DEPC
DK
DMEM
DMSO
DNA
dNTP
DSB
dsRNA
DTT
EBV
ECL
two-dimensional difference gel electrophoresis
three-dimensional
6-carboxyfluorescein
amino acid
4-(2-aminoethyl) benzenesulphonyl fluoride
Argonaute
alternate lengthening of telomeres
ampicillin resistance gene
anaphase-promoting complex
ammonium persulphate
alternate reading frame
adenosine triphosphate
American type culture collection
β-galactosidase
basic helix-loop-helix
Basic Local Alignment Search Tool
blasticidin resistance gene
base pair
bromo-deoxyuridine
bovine serum albumin
calcium chloride
cyclin-dependent kinase
cyclin-dependent kinase inhibitor
complementary deoxyribonucleic acid
colony forming units
chromatin immunoprecipitation
calf intestinal alkaline phosphatase
conserved region
carboxy-terminal
4'-6-diamidino-2-phenylindole
double-distilled
DNA damage response
diethyl pyrocarbonate
cyclin D1-CDK4R24C fusion construct
Dulbecco‘s modified Eagle medium
dimethyl sulphoxide
deoxyribonucleic acid
deoxyribonucleotide triphosphate
double strand break
double-stranded ribonucleic acid
dithiothreitol
Epstein Barr Virus
enhanced chemiluminescence
18
EcoR
EDTA
ES
FBS
FCS
FDR
G0
G1
G2
GFP
GPI
GSE
GTP
HAT
HCl
hCMV
HDAC
HDF
HEK293
HEPES
HGPS
HMF
HPV
hr
HRP
hTERT
IE
IL
IPTG
IR
IRES
kb
KCl
kDa
KOD
LB
Liquid N2
LOWESS
LT
LTR
M
M1
M2
mM
M phase
MCS
murine ecotropic receptor
ethylenediaminetetraacetic acid
embryonic stem
foetal bovine serum
foetal calf serum
false-discovery rate
quiescence
first gap phase
second gap phase
green fluorescent protein
glycosylphosphatidylinositol
genetic suppressor element
guanine triphosphate
histone acetylase
hydrochloric acid
human cytomegalovirus
histone deacetylase
human diploid fibroblast
human embryonic kidney 293
4-(2-hydroxyethyl)-1-piperazineethanesulphonic acid
Hutchinson-Gilford Progeria Syndrome
Human mammary fibroblast
Human Papilloma Virus
hour
horseradish peroxidase
catalytic component of human telomerase
immediate early promoter
interleukin
isopropyl--D-thiogalactopyranoside
ionizing radiation
internal ribosomal entry site
kilobase
potassium chloride
kilodalton
Thermococcus kodakaraensi DNA polymerase
luria broth base
liquid nitrogen
locally weighted linear regression
SV40 large T antigen
long terminal repeat
molar
mortality stage I
mortality stage II
millimolar
mitosis phase
multiple cloning site
19
MEF
MgCl2
MgSO4
min
miR
miRNA
micro-RNA
MoMuLV
MoMuSV
MOPS
MPF
mRNA
N
N2
NaCl
NaOH
NES
(NH4)2SO4
NHEJ
NKI
NLS
nt
N-terminus
OD
ORF
P
PAGE
PBS
PcG
PCR
PD
PMT
PNK
PTGS
puroR
R
REF
RIPA
RISC
RITS
RNA
RNAse
RNAi
ROS
rpm
RT
mouse embryo fibroblast
magnesium chloride
magnesium sulphate
minute
micro ribonucleic acid
micro ribonucleic acid
micro ribonucleic acid
Moloney Murine Leukaemia Virus
Moloney Murine Sarcoma Virus
3 – [N-morpholino] propanesulphonic acid
maturation promoting factor
messenger ribonucleic acid
ploidy
nitrogen
sodium chloride
sodium hydroxide
nuclear export signal
ammonium sulphate
non-homologous end-joining
Netherland Cancer Institute
nuclear localisation signal
nucleotide
amino-terminus
optical density
open reading frame
phosphorylation
polyacrylamide gel electrophoresis
phosphate buffered saline
polycomb group
polymerase chain reaction
population doubling
photomultiplier tube
polynucleotide kinase
post-translational gene silencing
puromycin resistance gene
restriction point
rat embryo fibroblast
radioimmunoprecipitation
ribonucleic acid -induced silencing complex
ribonucleic acid -induced transcriptional silencing
ribonucleic acid
ribonuclease
ribonucleic acid interference
reactive oxygen species
revolutions per minute
reverse transcription
20
RT-PCR
S phase
SA
SAHF
SAM
SA-β-Gal
SCF
SDS
sec
shRNA
SIPS
siRNA
SNP
STASIS
SV40
TAE
TEMED
TERC
TKO
T-OLA
ts
Ub
UNG
UTR
UV
UVP
V
v/v
w/v
WS
wt
X-gal
reverse transcription-polymerase chain reaction
deoxyribonucleic acid synthesis phase
senescence-associated
senescence-associated heterochromatic foci
significance analysis of microarrays
Senescence associated β galactosidase
Skp1/Cullin/F-box protein
sodium dodecyl sulphate
second
short hairpin ribonucleic acid
stress-induced premature senescence
short-interfering ribonucleic acid
single nucleotide polymorphism
stress or aberrant signalling-induced senescence
Simian Virus 40
Tris-acetate-EDTA
N,N,N‘,N‘ tetraethylenemethyldiamine
catalytic ribonucleic component of human telomerase
triple knockout
telomere oligonucleotide ligation assay
temperature sensitive
ubiquitination
uracil-N-glycosylase
untranslated region
ultraviolet radiation
dual intensity ultraviolet trans-illuminator
Volt
volume per volume
weight per volume
Werner Syndrome
wildtype
5-bromo-4-chloro-3-indolyl--D-galactopyranoside
21
ENCLOSED UNBOUND MATERIAL
Supplementary figures and tables are on a CD (in the cover).
STATEMENT CONCERNING COLLABORATIONS
The work presented in this thesis is the work of the author unless otherwise
indicated.
ACKNOWLEDGEMENTS
I would like to thanks Parmjit Jat for his fantastic support throughout my PhD. He has
been a precious help in guiding me through the various projects undertaken for this thesis
and helping me to link them all in one story. His patience for my ―frenchism‖ was most
appreciated.
Also, I owe many thanks to the great great people that I worked with in the Jat Lab: Tim
Szeto for his great humour in the lab, Nunu namely Parineeta Aurora for her consistent
support and friendliness, Kat namely Katharina Wanek for being my best chatting buddy
during the long hours in the tissue culture lab, Louise Mansfield for getting me started
with all the project and for being the most organized person I‘ve known so far, Annika
Alexopopoulo for being a model of dedication and work, and Mark, James and Catia, the
three undergrad students we received in our lab during that period, for being fresh,
interested and great fun.
I am also grateful to Mike O‘Hare, Anita Grigoradis, Ray Young and Holger Hummerich
for their helpful discussions, practical advice and participation in treating the data. A
special thank to Jess for being a great friend within the unit and keeping me in a good
mood whenever needed.
22
I would also like to extend my thanks to the Wellcome Trust to permit me financially to
undertake that big project.
For provision of reagents, I would like to thank Reuven Agami, Ole Gjoerup, Gregory
Hannon, Ed Harlow, Scott Lowe, Karl Munger, John Sedivy, Andrew Fry Leicester,
David Glover, Jesus Gil, Pascal Meier, Rene Medema, Xin Lu, Greg Towers and Didier
Trono, Andrei Gudkov, Julian Downward and Jay Morgenstern.
Finally, this project would not have been completed without the amazing support of both
my family and friends. So, I would like to thank Steven and Paula for their support during
the writing of the thesis. In particular, I would like to thank my parents, Christine and
Jean, Camille, my sister, and Chris, my boyfriend; I dedicate this work to them.
23
1
INTRODUCTION
1.1 REPLICATIVE SENESCENCE DISCOVERY
The first ―immortal‖ cells, discovered in a tissue culture laboratory, in 1951, were HeLa
cells, from the name of a patient, Mrs. Henrietta Lacks, from whom the biopsy was
extracted (Finkel, Serrano et al. 2007). This mother of five underwent a biopsy at John
Hopkins hospital for a suspicious cervical mass which was then identified as an
undifferentiated epidermoid carcinoma in the cervix. A portion of that biopsy also went
to George and Martha Gey‘s research laboratory. Unfortunately, Mrs. Lacks local lesion
could never be eradicated, and she died within six months of disseminated cancer. The
laboratory was more fortunate, however. This peculiar tumour grew very well in the
laboratory and because it could be transferred from generation to generation, it was
established as a perpetual cell line. At that point, scientists thought they had discovered
true immortal cells. They, however, quickly established that there was a limit to cell
division and that the cells would stop dividing and become specialised after a certain
number of divisions. Moorhead and Hayflick, more than 40 years ago, discovered that
normal human diploid fibroblasts stop dividing after 60-80 population doubling in
culture. In 1961, they proved that this growth arrest wasn‘t due to anything present in the
culture medium as they took early passage cells and transferred them into the
conditioned media without any changes but was due to some intrinsic factors, ―Hayflick
factors‖ (Hayflick and Moorhead 1961).These factors would accumulate inside the cells
until they senesced.
Today this proliferative limit named replicative is considered to be triggered largely by
erosion of the telomeres but also by various intrinsic and extrinsic factors such as DNA
damage, structure alteration, activation of certain oncogenes and physiological stress.
24
1.2 DEFINITION OF SENESCENCE
Cellular senescence was described for the first time in 1961 by Hayflick and Moorhead
as an irreversible growth arrest of human diploid cells that had lost their ability to divide
after a certain number of divisions (Figure 1.1) (Hayflick and Moorhead 1961).
Senescent cells are viable almost indefinitely, at least in vitro, even if they have stopped
dividing and synthesising DNA. They typically undergo dramatic morphological and
functional changes and acquire a very distinct gene and protein expression profile. For
instance, these cells acquire increased adhesion to the extracellular matrix and a
flattened and much enlarged phenotype with a vacuolated morphology (Chang, Broude
et al. 1999; Serrano and Blasco 2001; Narita, Nunez et al. 2003). A biochemical assay
has even been developed to detect senescent cells based on the increased senescenceassociated-β-galactosidase (SA-β-Gal) (Dimri, Lee et al. 1995; Shelton, Chang et al.
1999; Pascal, Debacq-Chainiaux et al. 2005). Confidence in the SA-β-Gal assay,
however, has been eroded by findings that its expression can be induced in some
immortalized cells and even reversed under some conditions (Herbig, Jobling et al.
2004). Another assay was developed as an alternative to test the senescence status using
three biomarkers (telomere dysfunction, activation of the ATM DNA-damage response,
and heterochromatinization of the nuclear genome).
Nevertheless, no precise link between these morphological and functional changes and
senescent signalling has been established so far and the senescence pathways outline has
yet to be defined.
1.3 TELOMERE INDUCED SENESCENCE
1.3.1 Telomeres
Telomeres are DNA-protein complexes at the ends of linear eukaryotic chromosomes.
Mammalian telomeres are composed of tandem repeats of a TTAGGG patterns of DNA
associated with proteins (Moyzis, Buckingham et al. 1988; Wellinger and Sen 1997).
25
Figure 1.1: Replicative Senescence
Hayflick and Moorhead [Hayflick and Moorhead, 1961] analysed primary HDFs sub-cultivated in vitro
and demonstrated that these cells exhibited a finite proliferative potential; after approximately 50 PDs, at a
point termed the ‗Hayflick limit‘, the cultures failed to expand, and the cells were considered senescent.
26
They were first discovered by Barbara McClintock and Herman Muller in the 1930‘s
and are capping structures that enclose and protect the ends of all eukaryotic linear
chromosomes from degradation (d'Adda di Fagagna, Teo et al. 2004).
The telomere induced senescence is the first molecular mechanism identified capable of
inducing irreversible cell growth arrest. The study of telomere length regulation revealed
that cells lose 50-200 base pairs of telomeric DNA with every single cell division during
S-phase and therefore progressively shortening their telomeres (Harley, Futcher et al.
1990; Blasco 2005). The human telomeres are around 8 to 12 kb at birth. Eventually,
telomeres reach a critical dysfunctional length that activates the p53 tumour suppressor
factor resulting in the cell senescence or apoptosis (de Lange 2005; von Zglinicki,
Saretzki et al. 2005). Only one or a few such telomere erosions are necessary to trigger
senescence (Martens, Chavez et al. 2000; Hemann, Rudolph et al. 2001).
1.3.2 Telomeres and DDR
In humans, several studies have shown a correlation between telomeres length, age and
aging diseases in a wide range of tissues (Cawthon, Smith et al. 2003; Panossian, Porter
et al. 2003; Ogami, Ikura et al. 2004; Canela, Vera et al. 2007). A large amount of
evidence demonstrated that telomere erosion was contributing to genome instability
(Maser and DePinho 2002) by initiating DNA damage response (DDR) (Figure 1.2).
However, mouse models with competent p53 pathways have recently shown that
telomere shortening could act as a tumour suppressor by promoting replicative
senescence (Figure 1.2). In opposition, in cells with mutant p53, the telomere induced
DDR triggers genome instability and tumourigenesis (Blasco, Lee et al. 1997; CosmeBlanco, Shen et al. 2007).
27
Figure 1.2: Telomeres shortening
In cells that do not express telomerase, telomeres become shorter with each cycle of cell division. a, Long
telomeres ensures that telomere ends, which are similar in chemical composition to broken DNA
sequences within chromosomes, are not mistaken for sites of DNA damage by the ATM- and ATRmediated DNA-repair machinery. b, When telomeres become critically short, they induce cellular
senescence. Such short telomeres were known to activate ATM and ATR kinases, which mediate the
DNA-damage response. Lazzerini Denchi and de Lange now identify structural changes that lead to the
activation of ATM and ATR at telomeres.
28
1.3.3 hTERT
Telomeres have a DNA damage repair system necessary for their maintenance through
the action of telomerases. Telomerase are ribonucleoproteins, with a catalytic DNA
polymerase activity called telomerase reverse transcriptase (TERT), which are in charge
of elongating telomeres (Greider and Blackburn 1985). There are two major components
of the telomerase holoenzyme: the telomerase reverse transcriptase (TERT) protein
subunit that catalyzes the enzymatic reaction of DNA synthesis and the telomerase RNA
(TR) component that serves as a template for the addition of deoxyribonucleotides to the
ends of chromosomes. The catalytic RNA is constitutive. TERT is generally turned off
in somatic cells. Although other proteins are associated with the holoenzyme, these two
components are essential and sufficient for telomerase activity and telomere lengthening
(Ishikawa 1997; Weinrich, Pruzan et al. 1997). However, most human adult tissues
express telomerase at levels not high enough to maintain the telomeres length intact and
this attrition results in aging (Collins and Mitchell 2002). The use of a telomerase
depleted mouse model helped to prove that the telomerase is the main cellular activity
responsible in the telomere maintenance (Blasco, Lee et al. 1997). This explains why
germ-line cells and cancer cells express TERT at high levels. Correspondingly, ectopic
expression of telomerase in vitro alone can contribute to the creation of immortalized
human fibroblast cell line from primary cells in particular cases (Bodnar, Ouellette et al.
1998).
Fundamentally, all human cancer cells have developed a mechanism to maintain
telomeres, essentially through an induction of telomerase activity (Stewart and Weinberg
2006). Alternatively, another mechanism exists, known as ALT for alternative
lengthening of telomeres, which involves inter telomeres homologous recombination
(Muntoni and Reddel 2005).
29
1.3.4 Telomerase and tumourigenesis
Consequently to these conclusions, the telomerase is often described as a tumourigenic
and an anti-aging factor. For example, it has been proven that mice deficient in
telomerase activity are cancer resistant while wild type mice would develop normally
tumours following various genetic alteration or carcinogenic treatments (GonzalezSuarez, Samper et al. 2000; Blasco 2005). These mice also display a shortened lifespan,
even from the first generation, which decreases with every new generation of deficient
mice (Blasco, Lee et al. 1997; Lee, Russo et al. 1998; Garcia-Cao, Garcia-Cao et al.
2006). Mice over-expressing telomerase, on the contrary, are prone to tumour
development (Gonzalez-Suarez, Samper et al. 2000; Canela, Martin-Caballero et al.
2004; Gonzalez-Suarez, Geserick et al. 2005) and an increased lifespan has been shown
in the few telomerase transgenic mice that do not develop cancer (Gonzalez-Suarez,
Geserick et al. 2005). However, it is important to note that telomerase induction cannot
prevent senescence caused by non-telomeric DNA damage or other inducers (Chen,
Prowse et al. 2001) as telomere shortening is only one of the causes of cellular
senescence.
1.4 DNA DAMAGE INITIATED SENESCENCE
All cells must protect their genomic integrity in order to guarantee a proper transfer of
the genetic information during the cell division. Cells respond to genotoxic stress
including DNA double strand breaks (DSBs) by activating a signalling cascade known
as DNA damage response (DDR). The DDR is a complex cascade of reactions regulated
by multiple and various DNA repair factors and cell cycle regulators, which seems to
converge on only one protein preferentially, p53, which is a key factor in the timely
execution of cell fate decisions. In addition, P53 is also downstream of telomere
shortening.
It is well established that DNA damage, especially DSBs, contribute to trigger
senescence (Di Leonardo, Linke et al. 1994; Parrinello, Samper et al. 2003) (Figure 1.3).
30
Figure 1.3: Causes and consequences of cellular senescence
Cellular senescence is triggered in response to a variety of intrinsic and extrinsic stimuli including
progressive shortening of telomeres, changes in telomeric structure at the ends of chromosomes or other
forms of genotoxic stress such as oncogene activation, DNA damage or oxidative stress resulting in a
DNA damage response and growth arrest via activation of the p53-p21 pathway (Ben-Porath and
Weinberg, 2004; Campisi and d‘Adda di Fagagna, 2007). When cellular senescence occurs, cellular
proliferation is lost, and the balance is tipped toward apoptosis and cell cycle arrest.
31
Recent data even suggest that DNA damage could be just a general common cause
underlying various different forms of cellular senescence such as oncogene-induced and
telomere-induced senescence (d'Adda di Fagagna, Reaper et al. 2003; Bartkova, Rezaei
et al. 2006; Di Micco, Amitrano et al. 2006) (Figure 1.3). In vitro cultured cells undergo
irreversible growth arrest when subject to various forms of DNA damage (te Poele,
Okorokov et al. 2002; Parrinello, Samper et al. 2003). The age-dependant accumulation
of DNA damage seems to be also a contributing factor to cellular senescence (Vijg
2000) therefore leading to an accumulation of senescent cells in aging tissues as well as
depletion in the number/function of stem cells.
1.5 ONCOGENE-INDUCED SENESCENCE
Oncogene-induced senescence (OIS) is a protective mechanism to avoid tumour
formation. The first human oncogene identified was Ras in 1982 and was found to be
able to transform immortalized rodent cells (Der, Krontiris et al. 1982; Parada, Tabin et
al. 1982) but needed additional DNA damage or genetic attrition to assist in
transforming primary cells (Land, Chen et al. 1986). In 1997, the accumulation of Ras in
wild type cells was proved to trigger proliferation followed by an irreversible growth
arrest accompanied by the accumulation of p53 and p16INK4A proteins (Serrano 1997).
This Ras-induced senescence was also found to be bypassed by the inactivation in vitro
of pRb and p53 pathways, suggesting similarities to tumour suppressor mechanisms.
The proof of oncogene-induced senescence has since then been demonstrated in vivo in
human tumour and mouse tumour models (Braig, Lee et al. 2005; Chen, Trotman et al.
2005; Collado, Gil et al. 2005; Michaloglou, Vredeveld et al. 2005; Courtois-Cox,
Genther Williams et al. 2006; Dankort, Filenova et al. 2007) (Figure 1.3). Furthermore,
mutations in K-ras, B-raf, PTEN and NF1 have been observed to trigger cellular
senescence in vivo. Senescence occurs in benign but not in advanced tumours,
supporting the first in vitro observation that activation of these pathways lead to an
initial burst of proliferation before causing cellular senescence (Courtois-Cox, Jones et
al. 2008).
32
1.6 CANCER AND SENESCENCE
Senescence can compromise tissue repair and regeneration and contribute to tissue and
organismal ageing due to depletion of stem/progenitor cell compartments. It could also
lead to removal of defective and potentially cancerous cells from the proliferating pool
thereby limiting tumour development (Campisi and d'Adda di Fagagna 2007; Collado,
Blasco et al. 2007) (Figure 1.3). In contrast to normal somatic cells, cancer cells have the
potential to proliferate indefinitely and this acquisition of an infinite proliferative
potential was proposed to be one of the six key events required for malignant
transformation (Hanahan and Weinberg 2000). The underlying mechanisms that control
cellular senescence, the signal transduction pathways involved and how the diverse
signals that result in senescence are all integrated remain poorly defined.
There are a lot of common key regulation checkpoints and very subtle differences
between tumourigenesis and senescence pathways and the balance between one another
is a fine line (Figure 1.3). For instance, both can be triggered, in different situations, by
DNA damage as results of DNA repair mechanisms activation (Bartkova, Rezaei et al.
2006; Halazonetis, Gorgoulis et al. 2008; Wang, Sengupta et al. 2008)
1.7 AGEING AND SENESCENCE
Several studies implicate a role for p53 and pRb in establishing senescence but also a
potential role as a regulator of organismal ageing (Tyner, Venkatachalam et al. 2002;
Maier, Gluba et al. 2004; Dumble, Moore et al. 2007). Although a physiological role for
p53 in ageing is controversial because it is supposed to extend lifespan by reducing the
occurrence of cancer, studies with different mouse models indicate a delicate balance
between tumour suppressive and age promoting functions of p53, under particular
circumstances. While pRb null mice are lethal, p53 null mice are viable but highly
cancer prone (Donehower, Harvey et al. 1992; Vooijs and Berns 1999).
33
On the other hand, Donehower has also described a mouse with a mutant p53 allele that
appears to enhance overall p53 activity, resulting in enhanced cancer resistance
accompanied by premature aging phenotypes and reduced longevity (Tyner,
Venkatachalam et al. 2002). Another lab also generated a p53 hypermorphic transgenic
mouse which displays even more dramatic accelerated aging (Maier, Gluba et al. 2004).
1.8 PATHWAYS OF SENESCENCE
Two pathways, p53 and pRb, are of particular importance concerning senescence and
have been (although never fully understood) extensively described in the literature. One
of the tasks this thesis is focusing on is to use the model to identify downstream
effectors of p53 and pRb and then investigate in normal cells whether they also are
relevant in these.
1.8.1 The p53 pathway
p53 is the quintessential tumour suppressor. p53, also named the ―guardian of the
genome‖ is primordial in maintaining the genomic integrity of the cells (Lane 1992;
Vogelstein, Lane et al. 2000). The importance of a functional p53 protein for a normal
cell cycle is emphasized by the fact that the p53 protein does not function correctly in
nearly half of all human cancers. In about half of these tumours, p53 is inactivated
directly as a result of mutations in the p53 gene. In many others, it is inactivated
indirectly through binding to viral proteins, or as a result of alterations in genes whose
products interact with p53 or transmit information to or from p53 (Vogelstein, Lane et
al. 2000). It has also been shown that p53-deficient mice show a very high incidence of
multiple, spontaneous tumours at an early age (Donehower, Harvey et al. 1992;
Donehower, Godley et al. 1995).
34
1.8.1.1 TP53 gene and p53 protein
P53 was first described in 1979 by its interaction with the viral protein SV40 LT antigen
and the adenovirus E1B 58K (DeLeo, Jay et al. 1979; Lane and Crawford 1979; Linzer
and Levine 1979). It was initially considered to be an oncogene but was subsequently
identified to be a tumour suppressor (Linzer and Levine 1979). The TP53 human gene is
located on chromosome 17 and possesses 11 exons whereas the mouse gene also
containing the same number of exons is situated on chromosome 11 (Soussi and May
1996). The human protein is ~53kDa (Hainaut, Soussi et al. 1997).
1.8.1.2 Functions of the p53 protein
The general assumption is that the p53 is normally present at low level partly as a result
of its degradation by the specific ubiquitin ligase MDM2, through the ubiquitinylationproteasome pathway but can be activated in cells as a response to various signals such as
DNA damage, stress, anoxia or depletion of the nucleotide pools. The tumour suppressor
ARF helps to stabilize p53 by binding and inhibiting MDM2. In response to stress
signals (perhaps the best studied of which is the response to DNA damage) p53 becomes
functionally active and triggers either a transient cell cycle arrest, cell death (apoptosis)
or permanent cell cycle arrest (senescence).
Both cellular senescence
(Sionov and Haupt 1999) and apoptosis (Heinrichs and
Deppert 2003) are potent tumour suppressor mechanisms that irreversibly prevent
damaged cells from going under neoplastic transformation. As a matter of fact, they also
were some of the first explored functions of p53. Later on, other important functions,
such as DNA repair (Albrechtsen, Dornreiter et al. 1999) and inhibition of angiogenesis
(Vogelstein, Lane et al. 2000), were discovered. p53 promotes longevity by reducing
somatic mutation and/or abnormal cell growth
and consequently reducing the
occurrence of cancer (Campisi 2003; Vijg, Busuttil et al. 2005). Recent evidence
suggests that an increased p53 activity can, at least under some circumstances, promote
35
organismal ageing (Tyner, Venkatachalam et al. 2002; Dumble, Gatza et al. 2004;
Maier, Gluba et al. 2004).
p53 is a sequence-specific transcription factor that binds to target consensus sites and
affects the transcription of its target genes (el-Deiry 1998). p53 regulates these genes
either by transcriptional activation (Murphy, Ahn et al. 1999) or by modulating other
protein activities by direct binding (Guimaraes and Hainaut 2002).
1.8.1.3 Regulation of p53 activity
The regulation of p53 activity can happen at various levels: p53 transcription, for
example, is effectively increased by DNA damage (Lu, Pochampally et al. 2000). It is
generally believed, though, that the principal mechanisms governing the activity of p53
occur at the protein level. These include post-translational modifications, regulation of
the stability of p53 protein, and control of its sub-cellular localization (Woods and
Vousden 2001). Of the post-translational modifications of p53, the most widely studied
and best-known so far is phosphorylation. After DNA damage induced by ionizing
radiation or UV light, phosphorylation takes place mostly at the N-terminal domain of
p53 (Appella and Anderson 2001). Another important modification is acetylation, which
(Ito, Adachi et al. 2001). In response to DNA damage, the p53 protein is also modified
by conjugation to SUMO-1, a ubiquitin-like protein (Gostissa, Hengstermann et al.
1999). Many proteins able to interact with p53 may also play a role in p53 regulation
(Vousden and Lu 2002).
Mdm2-mediated degradation regulates the stability of p53 (Figure 1.4). Mdm2 was
originally identified as a dominant transforming oncogene (Fakharzadeh, Trusko et al.
1991) and has been found to be amplified in human cancers (Momand, Jung et al. 1998).
Deletion of the mdm2 gene in mice is embryonically lethal, probably due to increased
accumulation of p53, but this lethality can be counter-acted by deletion of the TP53 gene
(Jones, Roe et al. 1995; Montes de Oca Luna, Wagner et al. 1995).
36
Figure 1.4: The p53 Signalling Pathway
Schematic diagram of the p53 signalling pathways that is involved in regulating progression through the
cell cycle in response to genotoxic stress or oncogenic signals. Ub: ubiquitin; P: phosphorylation.
37
The mdm2 protein regulates the activity of the p53 protein with many mechanisms such
as blocking the transcriptional activity of the p53 protein, exporting p53 from the
nucleus to the cytoplasm and promoting the degradation of p53 (Tao and Levine 1999;
Alarcon-Vargas and Ronai 2002). The p53-mdm2 relationship is vital in the regulation
of cell growth and death.
1.8.1.4 p19ARF protein
p19Arf (ARF, Alternative Reading Frame) is a protein capable of interacting with mdm2
(Kamijo, Weber et al. 1998) and interfering with the autoregulatory feedback loop
between the p53 and mdm2 proteins (Figure 1.5), thus increasing the amount of p53.
The gene that encodes the p19Arf protein also encodes p16INK4a. However, the p19Arf
protein is expressed by a separate promoter (Mao, Merlo et al. 1995). Both p16INK4a and
p19Arf are tumour suppressors (Zhang & Xiong 2001). The p19Arf protein is exclusively
localized in the nucleolus (Weber, Taylor et al. 1999) where it can bind to the central or
C-terminal portion of the mdm2 protein (Zhang, Xiong et al. 1998). There are currently
three competing theories about how p19Arf inhibits mdm2-mediated p53 degradation.
The first possibility is that the p19Arf protein sequesters the mdm2 protein into the
nucleolus, thus releasing p53 (Tao and Levine 1999; Weber, Taylor et al. 1999). The
second model suggests that nucleolar p19 ARF is relocalized by mdm2 to the nucleoplasm
and forms a ternary complex with mdm2 and p53, thus blocking the nuclear export of
both mdm2 and p53 (Zhang, Xiong et al. 1998). Additionally, p19Arf has been shown to
bind the p53 protein directly, indicating that it can, in addition to mdm2, recruit p53 into
ternary complexes (Kamijo, Weber et al. 1998). The third model proposes that, because
the p19Arf protein is able to bind to the mdm2 protein and inhibit its ubiquitin ligase
activity, p19Arf might prevent p53 nuclear export by blocking the ubiquitination of p53
(Honda and Yasuda 1999). It was shown by Weber and coworkers (Weber, Jeffers et al.
2000) that triple knock-out mice lacking functional p53, mdm2 and p19Arf proteins
develop tumours at a greater frequency than mice lacking p53 and mdm2 or p53 alone.
This suggests that p19Arf is a tumour suppressor independent of mdm2 and p53.
38
Figure 1.5:The pRb Signalling Pathway
Schematic diagram of the pRb signalling pathway involved in the cell cycle regulation in response to
genotoxic stress or oncogenic signals.
P: phosphorylation.
39
The p19Arf protein itself is regulated primarily at the transcriptional level. Both Myc and
E1A oncoproteins have been shown to induce the synthesis of p19Arf (de Stanchina,
McCurrach et al. 1998; Zindy, Eischen et al. 1998). In summary, these three proteins
form a system that regulates their localization, amount and function.
1.8.1.5 Oncogenic Ras
Mammalian ras genes are considered crucial in the regulation of cell proliferation (Bos
1989; Johnson, Greenbaum et al. 1997). In mammals, the Ras family consists of three
genes located on different chromosomes, encoding the homologous 21 kDa proteins HRas, N-Ras and K-Ras. It has been estimated that 30% of all human cancers express
mutated forms of ras (McMahon and Woods 2001). Ras can have either negative or
positive effects on cell growth, differentiation and death (Frame and Balmain 2000). The
signal is subsequently transmitted by a cascade of kinases, which results in the activation
of MAPK. The Ras-MAPK pathway is apparently involved in the regulation of basal
and induced levels of p53 (Fukasawa and Vande Woude 1997; Serrano, Lin et al. 1997).
In vascular smooth muscle cells, benzo(a)pyrene treatment has been shown to cause an
increase in Ras mRNA levels (Kerzee and Ramos 2000). Ras, in turn, induces p19Arf in
murine fibroblasts (Groth, Weber et al. 2000; Ferbeyre, de Stanchina et al. 2002). There
are also data that support a linear model from Ras through the induction of p19Arf to p53.
Palmero (Palmero, Pantoja et al. 1998) showed that an oncogenic form of Ras protein
increases significantly p19Arf mRNA. In ARF-/- mouse embryonic fibroblasts (MEF),
the p53 level is not affected by oncogenic Ras. In an earlier work on wild-type MEFs,
the p53 level increased after oncogenic Ras (Serrano, Lin et al. 1997). It can thus be
concluded that p19Arf is required for oncogenic Ras-induced accumulation of p53.
1.8.1.6
p21WAF1/Cip1/Sdi1
The p21CIP1/WAF1/Sdi1 protein was the first cyclin-dependent kinase inhibitor (CDKI)
identified (el-Deiry, Tokino et al. 1993; Harper, Adami et al. 1993; Noda, Ning et al.
1994). The p21CIP1/WAF1/Sdi1 protein has multiple functions. El-Deiry (1993) named this
40
gene WAF1 and found it to code a protein that mediates p53-induced growth arrest of
the cell cycle, and thus functions as a regulator of cell cycle progression at G1. Almost
simultaneously, another group showed it to be a regulator of CDK activity by its
interaction with a CDK (Harper, Adami et al. 1993). Yet another group demonstrated its
gene expression to be induced in relation to cellular senescence (Noda, Ning et al. 1994).
I has been shown that p21CIP1/WAF1/Sdi1 can inhibit all CDK-cyclin activities (Boulaire,
Fotedar et al. 2000), directly inhibit DNA replication (Li, Waga et al. 1994; Shivji, Grey
et al. 1994; Chen, Jackson et al. 1995) and at low level, act as an assembly factor for
CYD/CDK4,6 (LaBaer, Garrett et al. 1997; Cheng, Olivier et al. 1999).
The gene is transcriptionally up-regulated by wild-type p53 (el-Deiry, Tokino et al.
1993).The activation of p53 causes induction, directly downstream, of p21CIP1/WAF1/Sdi1,
which thanks to its promiscuous nature can, in turn, inhibit all CDK-cyclin complexes
and arrests the cell at different stages of the cell cycle (Gartel, Serfas et al. 1996;
Colman, Afshari et al. 2000; Taylor and Stark 2001) (Figure 1.4). This gives time for
DNA repair before replication or mitosis and thus links p21CIP1/WAF1/Sdi1 directly to the
tumour suppressor function of p53.
1.8.1.7 p53 family: p63 and p73 proteins
Two genes notably similar to the TP53 gene seem to be of importance in the cell cycle.
One of these genes is called p63, p51 or KET, (Schmale and Bamberger 1997; Osada,
Ohba et al. 1998; Yang, Kaghad et al. 1998) and the other p73 (Kaghad, Bonnet et al.
1997). They encode proteins that share high sequence similarity and conserved
functional domains with p53 and can exert p53-like functions, such as transactivation of
p53 target genes and induction of apoptosis (Yang, Kaghad et al. 2002). Both give rise
to differentially spliced mRNAs and then, respectively, to several different proteins
homologous to p53 (Levrero, De Laurenzi et al. 2000). There are at least three different
forms of the p63 protein differing at the C-terminal end (α, β and γ) that may also differ
within the transactivation domain (p63TA and p63ΔDN) and six different variants of the
p73 protein, p73-. The p73 protein, like p53, accumulates in response to DNA damage,
41
and it is noteworthy that different types of inducers of DNA damage seem to affect p73
in different ways(Levrero, De Laurenzi et al. 2000). Both p63 and p73 take part in the
regulation of normal cell development and apoptosis (Lohrum and Vousden 2000).
Different forms of p63 protein can act in a dominant-negative manner towards p53
(Yang, Kaghad et al. 1998), but whether p63 dysregulation has a role in tumourigenesis
remains to be seen. p73, on the other hand, has been suggested to be a tumour suppressor
protein(Levrero, De Laurenzi et al. 2000), although opposite opinions have also been
presented (Irwin and Kaelin 2001). The function of p63 or p73 as a tumour suppressor
still remains unclear (Michael and Oren 2002).
1.8.2 The pRb pathway
1.8.2.1 Cell cycle, cyclins and CDKs
Senescence is by definition an irreversible arrest of the cell cycle; therefore, it is no
surprise that cell cycle and senescence share an intricate web of their respective
pathways. Cyclins were the first discovered cell cycle regulators, their expression levels
increasing before mitosis and decreasing during cytokinesis (Evans, Rosenthal et al.
1983). They are divided into different category each sporting a specific role in the cell
cycle sequence. Cyclins A have been associated with Mitosis and the S-phase (DNA
synthesis phase) of the cycle whereas cyclins B were only associated with the mitosis
and cyclins E with the S-phase. Cyclins D were linked to the G1-phase (Roberts 1999).
Cyclins function by activating cyclin-dependant kinases (CDK) through binding. These
CDKs, in opposition, conserve stable expression levels throughout all the cell cycle.
1.8.2.2 CDK inhibitors
The Cyclin-CDK activity is also regulated by CDK inhibitors (CDKIs). These CDKI
have proven to be of great importance and have been classified into 2 families, namely
the INK4A family and the Cip/Kip family.
42
INK4A family
The INK4A family consists of p16INK4a, p15INK4b, p18INK4c and p19INK4d. INK4A family
members function by inhibiting the kinase activity of CDK4 and CDK6 (Serrano,
Hannon et al. 1993) (Figure 1.5). p16INK4a and p15INK4b are known to be associated with
tumour suppression while p18INK4c and p19INK4d are highly expressed during
development (Zindy, Soares et al. 1997).
Cip/Kip family
The Cip/Kip family consists of family members p21CIP1/WAF1/Sdi1, p27 and p57 (Sherr and
Roberts 1999). All the members of this family bind and inactivate CDK2 complexes;
however, the mechanism by which they inactivate the complexes varies between them.
The Cip/Kip family also functions as both positive and negative regulators of the
CDK4/6 complexes; p21CIP1/WAF1/Sdi1, for example, acts as an assembly factor for
CDK4/6 complexes at low levels but turns into an inactivator while its levels increase.
1.8.2.3 CDKs and E2F
During G0 (quiescence) and early G1 (first gap phase) of the cell cycle, a combination
of low levels of cyclins and high CDKI activity ensures pRb remains bound to the E2F
transcription factor (Figure 1.5). Then, in response to extracellular signals, such as
mitogens, D-type cyclins start to accumulate and to increase the cyclin D-CDK4/6
activity. This results in the phosphorylation pRb and the subsequent release of E2F. This
permits transcriptional activation of E2F-responsive genes required for S-phase
(Weinberg 1995; Bartek, Bartkova et al. 1996).
1.8.2.4 Rb family of proteins
One of the major targets of the cyclin-CDK kinases is the Rb family of proteins. The Rb
family is defined by the possession of a bipartite pocket region and is comprised of three
members, pRb, p107 and p130. The pocket region consists of two conserved domains
that are separated by a spacer region. In the case of pRb, this region encompasses aa
43
379-928 (Lee, Shew et al. 1987; Hannon, Demetrick et al. 1993; Mayol, Grana et al.
1993; Zhu, van den Heuvel et al. 1993).
The fact that Rb family members exhibit a high level of sequence homology results in
some level of functional redundancy; for example, Rb family members share common
activities in the regulation of cell proliferation, differentiation and apoptosis (Claudio,
De Luca et al. 1996). However, functional specificity of individual family members has
also been described; for example, p107 is the predominant family member that remains
bound to E2F-responsive promoters; p130 is the predominant family member that
remains bound during G1 phase (Takahashi, Rayman et al. 2000; Rayman, Takahashi et
al. 2002), whereas pRb is commonly expressed in both proliferating and nonproliferating cells.
1.8.2.5 pRb gene
The RB1 gene encodes pRb, a ubiquitously expressed 105 kilodalton (kDa) protein. The
fact that pRb is ubiquitously expressed and regulated in a cell cycle-dependent manner is
consistent with it functioning as a general regulator of the cell cycle (Lee, Shew et al.
1987; Buchkovich, Duffy et al. 1989; DeCaprio, Ludlow et al. 1989; Cobrinik, Dowdy
et al. 1992). Its primary function is to inactivate the E2F family of transcription factors
during G1/S phase transition of the cell cycle, yet over 100 other pRb-binding proteins
have been described (Morris and Dyson 2001). These include cell cycle regulated
proteins such as Mdm2 (Xiao, Chen et al. 1995), PML (Alcalay, Tomassoni et al. 1998)
or helix-loop-helix proteins involved in differentiation (Iavarone, Garg et al. 1994;
Alani, Young et al. 2001).
1.8.2.6 pRb discovery
RB1 was the first tumour suppressor gene to be cloned in humans and originally formed
the basis of Knudson‘s two-hit hypothesis (Knudson 1971), a hypothesis that was
supported by evidence derived from analysis of patients with hereditary and non44
hereditary forms of retinoblastoma, a rare tumour of the eye; Knudson showed that
individuals with the hereditary form of retinoblastoma often developed bilateral tumours
whereas individuals with the non-hereditary form usually developed unilateral tumours.
This led Knudson to hypothesize that two mutational events were required to inactivate
the gene responsible for retinoblastoma, but, in individuals that inherited a mutation in
the retinoblastoma gene, only one mutational event was required to inactivate the
remaining functional allele. RB1 was subsequently identified as the gene that was causal
to this process and it has since been shown that most human cancers harbour mutations
that directly or indirectly compromise pRb function (Murphree and Benedict 1984;
Sellers and Kaelin 1997); as an example, inactivating mutations frequently occur in RB1
itself, in addition to the mutation of upstream regulators of pRb, such as the homozygous
deletion of p16INK4a or amplification of the CDK4 locus. Significantly, most tumourassociated RB1 mutations occur in the pocket protein domain (Hu, Dyson et al. 1990;
Huang, Wang et al. 1990; Classon and Dyson 2001).
pRb was also found to be sequestered and thereby inactivated by SV40 LT antigen,
Polyoma LT antigen, Adenovirus E1A protein, HPV 16/18 E7 protein (DeCaprio,
Ludlow et al. 1988; Dyson and Harlow 1992; Moran 1993; Mymryk and Bayley 1994;
Eckner, Ludlow et al. 1996).
1.8.2.7 pRb function
During G0 and early G1, the C-terminal domain of pRb is hypophosphorylated (Knudsen
and Wang 1996; Bonetto, Fanciulli et al. 1999). This enables pRb to bind directly to and
inactivate (Figure 1.5) E2F in two ways; firstly, by binding to an 18 amino acid motif
within the E2F transactivation domain, pRb directly blocks the ability of E2F to form a
transcriptionally active complex (Flemington, Speck et al. 1993; Helin, Harlow et al.
1993). Secondly, pRb recruits repressive complexes such as histone deacetylase
(HDAC) complexes and histone methyltransferases to the promoter regions of these
genes to actively repress E2F transcription (Brehm, Miska et al. 1998; Luo, Postigo et
al. 1998; Zhang, Postigo et al. 1999; Chen and Wang 2000; Dahiya, Gavin et al. 2000;
He, Cook et al. 2000; Lai, Kennedy et al. 2001; Frolov and Dyson 2004). pRb also binds
to a heterochromatic protein, HP1, via its LXCXE motif to promote the binding of HP1
45
to modified histones. HP1 uses its chromodomain to directly bind to modified histones,
in addition to adjacent histone tails, thereby spreading the transcriptional silencing signal
to nearby nucleosomes (Bannister, Zegerman et al. 2001; Lachner, O'Carroll et al. 2001;
Nielsen, Schneider et al. 2001). This activity leads to the formation of a compact DNA
structure that is inaccessible to transcription factors.
Evidence to support the role of pRb in transcriptional silencing includes the fact that,
during G1 phase, pocket proteins can be detected in peri-nucleolar foci that also contain
E2Fs and histone desacetylases (Kennedy, Barbie et al. 2000).
During mid G1 phase, pRb is phosphorylated by the activity of cyclin D1-CDK4/6
(Figure 1.5). At R and late G1 phase, pRb is further phosphorylated by the activity of
cyclin E-CDK2 (Adams 2001) (Figure 1.5). Hyperphosphorylation of pRb in the Cterminal domain peaks during late G1 phase and causes pRb to dissociate from E2F
(Weinberg 1995; Knudsen and Wang 1996; Bonetto, Fanciulli et al. 1999). This is
supported by evidence that loss of Rb family repressor complexes at E2F-responsive
promoters enables E2F to induce expression of S phase genes required for DNA
synthesis (Takahashi, Rayman et al. 2000; Rayman, Takahashi et al. 2002; Taubert,
Gorrini et al. 2004). pRb is maintained in its hyperphosphorylated form until emergence
from M phase (Weinberg 1995), when it is dephosphorylated by PP1, a type 1
serine/threonine phosphatase (Nelson, Krucher et al. 1997).
In addition to regulating cell cycle progression, pRb also plays a role in senescence,
differentiation and apoptosis. During senescence, pRb interacts with HP1 and histone
methyltransferases such as SUV39H1 to specifically repress E2F-responsive promoters
and maintain the senescent state. However, experimental data is limiting due to the
difficulty of obtaining good immunofluorescence data from compact chromatin (Narita,
Nunez et al. 2003). In contrast, differentiation requires the direct interaction of pRb with
tissue specific transcription factors to induce the differentiation of many different cell
lineages, including adipogenesis, myogenesis and haematopoiesis (Gu, Schneider et al.
1993; Dunaief, Strober et al. 1994; Condorelli, Testa et al. 1995; Condorelli and
46
Giordano 1997). pRb activity is essential for this process (Lee, Chang et al. 1992), and
this is shown by the inability of cells from mice deficient in pRb activity to differentiate
both in vitro and in vivo (Maione, Fimia et al. 1994; Slack, Skerjanc et al. 1995; Lipinski
and Jacks 1999; Thomas, Carty et al. 2001; Classon and Harlow 2002; de Bruin, Maiti et
al. 2003). There is also some evidence to indicate that pRb can inhibit apoptosis; for
example, reconstitution of pRb in Saos-2 cells (a p53- and pRb- null osteosarcoma cell
line) is sufficient to bypass apoptosis induced by exposure to ionizing radiation (IR)
(Haas-Kogan, Kogan et al. 1995). Moreover, functional pRb activity is sufficient to
inhibit IFN -induced apoptosis (Berry, Lu et al. 1996).
1.8.2.8 E2F
The predominant function of Rb family members is to negatively regulate E2F activity.
Consequently, E2F plays a critical role in the cell cycle regulation and this is shown by
the fact that E2F activity is commonly abrogated during tumourigenesis; for example,
deregulation of the E2F family occurs in almost all cancers (Phillips and Vousden 2001),
whereas over-expression of E2F1 induces senescence in primary HDFs (Dimri, Itahana
et al. 2000).
E2F functions as a transcriptional regulator by forming a heterodimer with its cognate
partner DP. Two DP proteins have been identified, namely DP1 and DP2, their
heterodimerisation enhances both E2F transactivational activity, and the ability of Rb
family members to bind to and negatively regulate E2F. Seven E2F family members
have been described to date and these can be sub-divided into transcriptional activators
(E2Fs 1-3a) and transcriptional repressors (E2Fs 3b-7). The lack of transactivation and
pocket protein-binding domains in E2F6 (Cartwright, Muller et al. 1998; Gaubatz, Wood
et al. 1998; Trimarchi, Fairchild et al. 1998) is thought to render this particular E2F as a
repressor as it prevents activator E2Fs from binding to the DNA and/or recruits
polycomb group (PcG) proteins to target genes (Trimarchi, Fairchild et al. 2001). E2F7
represents a recently identified E2F family member that is likely to function as a
transcriptional repressor, as determined by sequence analysis (de Bruin, Maiti et al.
2003).
47
1.8.3 Common pathways
1.8.3.1 INK4A Locus
The INK4A locus situated on human chromosome 9p21 is amongst the most frequent
sites of genetic loss in human cancer and constitutes a unique feature in eukaryotes in
the fact that it results in two splice variants that both encode tumour suppressor proteins,
namely p16INK4a and p14ARF. These proteins share no sequence homology at the protein
level and differ in their functional activity, yet both function to negatively regulate
distinct pathways that are critical for cell cycle progression: p16INK4a regulates the pRb
pathway whereas p14ARF regulates the p53 pathway. Both p16INK4a and p14ARF share
common regulatory mechanisms since they are both induced in response to aberrant
growth or oncogenic stress, and both can be induced upon senescence. Yet, whilst there
is substantial evidence to associate functional inactivation of p16 INK4a with
tumourigenesis, evidence to link p14 ARF inactivation to tumourigenesis is less clear. This
is due to the fact that p14ARF promoter methylation and missense mutations specific to
p14ARF are rare and p14 ARF has not been as extensively analysed as p16INK4a in the
context of human cancer.
Moreover, p14ARF activity is often lost concomitantly with p16INK4a and/or p15INK4b; for
example, p15INK4b is located only 10 Kilobases (kb) from the first exon of p14ARF,
therefore, co-deletion of p14ARF with p15INK4b frequently occurs. It is likely that p16INK4a
and p14ARF evolved by selection of a common function and this hypothesis is supported
by their common ability to function as tumour suppressive proteins, their ability to be
expressed under similar conditions and co-regulated by molecules such as Bmi-1, CBX7
and TBX2 (Jacobs, Kieboom et al. 1999; Jacobs, Keblusek et al. 2000; Gil, Bernard et
al. 2004).
48
1.8.3.2 p16INK4a
The role of p16INK4a as a tumour suppressor was first indicated by studies of familial
melanoma that showed that incidences of melanoma segregated with missense mutations
in p16INK4a (Hussussian, Struewing et al. 1994; Holland, Beaton et al. 1995; Liu, Lassam
et al. 1995; Zuo, Weger et al. 1996). It has since been shown that p16INK4a is inactivated
by deletion, point mutation and promoter methylation in many primary tumours and
derived cell lines. However, the fact that humans homozygous for a severely truncated
form of p16INK4a may remain tumour free for several decades indicates that loss of
p16INK4a activity is not sufficient to induce tumourigenesis. More likely, p16INK4a
cooperates with other events (Gruis, Weaver-Feldhaus et al. 1995; Pavel, Smit et al.
2003).
P16INK4a functions by specifically inactivating cyclin D-containing CDK complexes;
p16INK4a binds to and induces a conformational change in CDK4/CDK6 that results in
the inhibition of adenosine triphosphate (ATP) -binding and thereby disrupts the
interaction with D-type cyclins. This activity prevents CDK4/6 from phosphorylating
pRb (Alcorta, Xiong et al. 1996; Hara, Smith et al. 1996; Serrano 1997; Kiyono, Foster
et al. 1998; Zhu, Woods et al. 1998; Ohtani, Zebedee et al. 2001; Schmitt, Fridman et al.
2002). Evidence to support this includes the fact that loss of pRb and p16INK4a activity
generally occurs as mutually exclusive events in non-small cell lung cancer (Otterson,
Kratzke et al. 1994; Shapiro, Park et al. 1995). Moreover, p16INK4a expression cannot
efficiently arrest pRb-deficient cell lines (Lukas, Parry et al. 1995).
p16INK4a is positively regulated at the transcriptional level by Ets-1, a transcriptional
activator that is activated by phosphorylation via ERK and p38 in response to Ras
signalling. This pathway is subject to negative regulation; for example Wip-1
phosphatase negatively regulates p38 (Bulavin, Phillips et al. 2004) and other negative
regulators of p16INK4a include Bmi-1 (B lymphoma Moloney Murine Leukaemia Virus
(MoMuLV) insertion region 1; (Itahana, Zou et al. 2003; Park, Morrison et al. 2004),
and Id1 (Inhibitor of DNA-binding 1) (Zheng, Wang et al. 2004).
49
P16INK4a also represents one of the few genes up-regulated upon replicative senescence
and maintained at a high level in senescent cells. This up-regulation correlates with the
constitutive hypo-phosphorylation of pRb in senescent cells (Alcorta, Xiong et al. 1996;
Hara, Smith et al. 1996; Zindy, Soares et al. 1997; Stein, Drullinger et al. 1999). In
contrast, inactivation of p16INK4a is sufficient to enable some human cell types to become
immortalised in conjuction with reconstitution of telomerase activity; for example,
human mammary epithelial cells and keratinocytes (Kiyono, Foster et al. 1998;
Rheinwald, Hahn et al. 2002).
1.8.3.3 p14ARF
p14ARF was originally identified as a splice variant of the INK4A locus and is also known
as ARF (Alternate Reading Frame), p14ARF in humans, or p19Arf in mice. p14ARF has its
own promoter and differs to p16INK4a by the inclusion of an alternative first exon
(Quelle, Zindy et al. 1995). This results in the translation of p14ARF in an alternate
reading frame to p16INK4a, so that it exhibits no amino acid homology to p16INK4a. The
first indication that p14ARF functioned as a tumour suppressor came from the observation
that mice lacking the first exon of p14ARF were prone to spontaneous and carcinogeninduced tumours (Serrano, Lee et al. 1996). Loss of p14ARF activity was subsequently
shown to render p53 inactivation surplus for immortalisation of MEFs, both in vitro
(Kamijo, Zindy et al. 1997) and in tumours in vivo (Chin, Pomerantz et al. 1997), and
could inhibit transformation of MEFs by Mdm2. Yet, this activity did not occur in cells
lacking p53 (Pomerantz, Schreiber-Agus et al. 1998). This indicated that p14ARF
functioned upstream of p53 in a linear pathway. p14ARF specific mutations have since
been reported in incidences of familial melanoma and astocytoma (Randerson-Moor et
al, 2001; Rizos et al, 2001). Moreover, promoter methylation of p14ARF, but not p16INK4a
was implicated in some incidences of colon cancer (Esteller, Tortola et al. 2000;
Esteller, Gonzalez et al. 2001; Sato, Harpaz et al. 2002), and the finding that TBX2 and
Pokemon, two transcriptional repressors of p14ARF (Jacobs, Keblusek et al. 2000;
Maeda, Hobbs et al. 2005), are both aberrantly over-expressed in a subset of human
breast cancers and lymphomas indirectly links p14ARF to human cancer.
50
p14ARF functions by sequestering Mdm2 to the nucleolus, thereby impairing the ability
of Mdm2 activity to promote the degradation of p53 by ubiquitin-mediated proteolysis
(Pomerantz, Schreiber-Agus et al. 1998; Zhang, Xiong et al. 1998; Sherr 2000; Sherr
and DePinho 2000). This activity enables p14 ARF to indirectly stabilise p53 (Weber,
Taylor et al. 1999; Sherr and Weber 2000; Lowe and Sherr 2003). The N-terminal 25 aa
are critical for p14ARF functional activity, and this region is encoded entirely by exon 1
(Quelle, Zindy et al. 1995). It has also been shown that p14ARF can inhibit cell
proliferation by p53-independent pathways (Cleveland and Sherr 2004)
p14ARF expression is repressed under normal cellular conditions but is activated in
response to aberrant signalling; for example, in response to oncogenic signals such as cMyc, E2F-1, oncogenic Ras, v-abl, DMP1 and -Catenin (DeGregori, Leone et al. 1997;
Dimri, Itahana et al. 2000; Inoue, Wen et al. 2000; Inoue, Zindy et al. 2001; Sherr 2001).
The p19Arf-p53 pathway is the major pathway that induces senescence in mice (Lowe
and Sherr 2003; Sharpless and DePinho 2005) since p19Arf expression correlates with the
onset of senescence in MEFs and since cells that lack p19 Arf do not senesce in culture
(Kamijo, Zindy et al. 1997; Zindy, Soares et al. 1997). Moreover, p19Arf-null mice are
prone to develop spontaneous tumours (Kamijo, Zindy et al. 1997; Kamijo, Bodner et al.
1999). There is also evidence to suggest that over-expression of E2F1 induces p14ARF,
thereby negatively regulating Mdm2 activity and stabilising p53 (DeGregori, Leone et
al. 1997; Prives 1998; Sherr and DePinho 2000). This activity indirectly links the pRb
and p53 pathway, and also links E2F activity to the induction of senescence (Zhu,
Woods et al. 1998). However, the significance of this pathway in humans is unclear; for
example, despite the fact that p14ARF over-expression can induce cell cycle arrest or
senescence (Quelle, Zindy et al. 1995; Kamijo, Zindy et al. 1997; Dimri, Itahana et al.
2000; Wei, Hemmer et al. 2001), it has been argued that p14ARF activity is not critical
for these processes (Munro, Stott et al. 1999; Wei, Hemmer et al. 2001; Rheinwald,
Hahn et al. 2002; Sharpless and DePinho 2005). Moreover, p14ARF expression levels rise
only in some HDF strains upon replicative senescence (Dimri, Itahana et al. 2000).
51
1.9 DNA TUMOUR VIRUSES
SV40, Adenovirus, and Human Papilloma Virus (HPV) are three examples of DNA
tumour viruses. The natural hosts of DNA tumour viruses are differentiated cells,
therefore, these viruses have evolved mechanisms to enable them to replicate in a nonproliferative cellular environment. The mitogenic properties include the ability to alter
the cellular transcription machinery to promote the expression of proteins that are
required for viral replication, to overcome the finite proliferative potential and to block
cellular defences against viral intrusion. Consequently, some of the viral proteins
encoded by the DNA tumour viruses are able to inactivate the major control pathways
regulating the cell cycle and are therefore implicated in the induction of tumourigenesis;
for example, SV40 LT, HPV Type 16 E6 and E7 and Adenovirus Type 5 E1A and E1B
all function as potent viral oncoproteins to induce immortalisation and transformation of
many cell types (Braithwaite, Cheetham et al. 1983; Caporossi and Bacchetti 1990;
Chang, Ray et al. 1997; Duensing and Munger 2002). This has led to the extensive use
of these viruses as molecular tools to delineate many signalling pathways in mammals.
Importantly, these viral oncoproteins were the first to reveal the critical roles of p53 and
pRb in the regulation of the cell cycle.
1.9.1 SV40
SV40 is a member of the papovavirus family of small icosahedral DNA viruses. SV40
was first linked to tumourigenesis by its ability to stably transform a proportion of
hamster and rodent cell lines infected with this virus. Infection of newborn hamsters
with SV40 induced the formation of tumours (Hilleman 1998). Unlike the natural lytic
lifecycle of SV40 in its natural hosts (rhesus monkey or African green monkey cells),
human or hamster cells are semi-permissive to infection with SV40; infection of these
cells is sufficient for early SV40 genes to be expressed in a transient manner and survive
infection. Moreover, a small proportion of infected cells permit viral replication. In
contrast, mouse cells can be infected with SV40 but are non-permissive for viral
replication and do not produce progeny virus particles.
52
1.9.2
LT
Three antigens are expressed from the SV40 early region by differential splicing of the
same messenger RNA (mRNA) transcript; namely, large T (LT) antigen, small t antigen
and 17 kT antigen. The 708 aa LT protein alone is responsible for many of the functions
of SV40 that are required for it to complete its lifecycle. LT is also involved in
promoting the immortalisation of many cell types; for example, LT activity is sufficient
to bypass replicative senescence in rat embryo fibroblasts (Jat and Sharp 1989).
Moreover, LT activity is required to maintain these cells in an immortalised state since
inactivation of LT results in a rapid and irreversible arrest in either G 1 or G2 phase (Jat
and Sharp 1989). This indicates that the endogenous senescence machinery remains
intact during this process. In accordance with this finding, MEFs become dependent
upon LT for maintaining growth only when their normal mitotic lifespan has elapsed
(Ikram, Norton et al. 1994).
LT possesses multifunctional activity; for example, it possesses both DNA and RNA
helicase activity (Scheffner, Knippers et al. 1989), ATPase activity (Tjian and Robbins
1979), RNA-binding activity (Carroll, Samad et al. 1988), DNA-binding activity
(Carroll, Hager et al. 1974) and transcriptional regulation activity (Alwine, Reed et al.
1977; Hansen, Tenen et al. 1981; Mitchell, Wang et al. 1987; Saffer, Jackson et al. 1990;
Zhu, Rice et al. 1991; Gilinger and Alwine 1993; Gruda, Zabolotny et al. 1993; Rice and
Cole 1993; Rushton, Jiang et al. 1997). LT can also impair the activities of many host
cell proteins such as p53 (Lane and Crawford 1979; Linzer and Levine 1979), pRb
(DeCaprio, Ludlow et al. 1988) , p107 (Dyson, Buchkovich et al. 1989; Ewen, Ludlow
et al. 1989), p130 (Hannon, Demetrick et al. 1993), CBP, BUB1 (Cotsiki, Lock et al.
2004; Williams, Roberts et al. 2007), p300 (Avantaggiati, Carbone et al. 1996; Eckner,
Ludlow et al. 1996) and TBP (Martin, Subler et al. 1993). Nuclear localisation is
mediated the N-terminal region of LT (Soule and Butel 1979; Kalderon, Richardson et
al. 1984).
LT shares significant sequence homology to the conserved region 2 (CR2) domain of
E1A and E7 protein between amino acid residues 103-107. However, within SV40 LT,
53
there is a CR2 domain that has LXCXE which can be functionally swapped. This region
contains the canonical LXCXE-binding motif that mediates stable Rb family binding
(DeCaprio, Ludlow et al. 1988; Moran 1988; Munger, Werness et al. 1989). LT only
binds to the hypophosphorylated and therefore active form of Rb family members
(Ludlow, DeCaprio et al. 1989; Ludlow, Shon et al. 1990). Consequently, LT promotes
the release of E2F, enabling it to activate transcription from E2F-responsive promoters.
This activity is critical for immortalisation as mutants defective for pRb-binding exhibit
a reduced ability to immortalise rodent cells (DeCaprio, Ludlow et al. 1988; Powell,
Darmon et al. 1999). There is evidence to suggest that pRb-binding is important for the
ability of LT to transform cells since some pRb-binding LT mutants are defective for
transformation (Ali and DeCaprio 2001).
P53 was originally identified as a LT-binding protein (Lane and Crawford 1979; Linzer
and Levine 1979) and binding to p53 is mediated by a bipartite region located towards
the C-terminus of the protein between amino acid residues 351-450 and 533-626. LT
interaction with p53 occurs via direct binding of LT to the sequence-specific DNAbinding domain of p53, as mutants of p53 that are impaired in sequence-specific DNAbinding activity are unable to bind to LT. The interaction of LT with p53 leads to
abrogation of p53 activity since p53 is unable to transcriptionally regulate its target
genes. This interaction stabilises p53 as both the half-life and steady-state levels of p53
are increased (Oren and Levine 1981; Deppert, Haug et al. 1987). It has also been
suggested that the association of p300 and Mdm2 with p53 in a LT-binding complex
contributes to this activity (Brown, Deb et al. 1993; Henning, Rohaly et al. 1997;
Grossman, Perez et al. 1998). The ability of LT to impair p53 activity appears to be
critical for the immortalisation of MEFs (Conzen and Cole 1995). This is in contrast to
data derived from rat embryo fibroblasts since LT mutants that lack the C-terminal p53
bipartite binding domain are able to immortalise (Powell, Darmon et al. 1999). This
indicates that additional activities of LT may be able to inactivate downstream effectors
of p53, and this may be mediated via Rb family binding (Quartin, Cole et al. 1994;
Rushton, Jiang et al. 1997). P300 and CBP binding sites are also present in both the N-
54
terminal and C-terminal domains of LT (Eckner 1996; Lill, Grossman et al. 1997),
although their interactions may occur indirectly via p53-binding.
1.9.3 Adenovirus Type 5
Adenoviridae are double-stranded DNA viruses, 51 different serotypes have been
identified. They primarily infect host epithelial tissues in the lung or enteric system and
have been associated with the development of acute respiratory diseases. Adenovirus
type 12 was the first serotype to be identified as being associated with tumourigenesis in
rodents (Trentin, Yabe et al. 1962), but there is no evidence to indicate that adenovirus
can induce tumourigenesis in humans. Transcription of the adenovirus genome is
regulated by virus-encoded regulatory factors and two of the genes to be transcribed are
E1A and E1B.
1.9.3.1 E1A
E1A represents a major regulatory protein expressed very early during adenovirus
infection that is capable of activating transcription from a variety of viral and cellular
promoters and notably all the other genes encoded within the viral genome. Like LT,
E1A exhibits multifunctional activity and can directly bind to multiple cellular proteins
required for cell proliferation to mediate this activity. Indeed, Rb family members,
cyclin A, p300 and others were originally identified by their interaction with E1A
(Whyte, Buchkovich et al. 1988; Faha, Ewen et al. 1992). E1A is synthesised almost
immediately after infection and two of the most abundant products are the 13S and 12S
E1A splice variants (Perricaudet, le Moullec et al. 1980).
The conserved CR2 motif defines the region in E1A where Rb family members bind
(Harlow, Whyte et al. 1986; Whyte, Buchkovich et al. 1988; Whyte, Williamson et al.
1989). However, residues in conserved region 1 (CR1) of E1A are also involved in this
process (Whyte, Williamson et al. 1989). This interaction disrupts pRb-E2F complexes
and enables E2F to promote entry into S phase (Sherr 1996). E1A is localised to the
nucleus by virtue of a highly basic pentapeptide signal sequence located at the extreme
55
C-terminus. E1A, like LT, can also bind to p300 (Dorsman, Hagmeyer et al. 1995;
Wang, Moran et al. 1995; Goodman and Smolik 2000), and this promotes the formation
of a pRb/p300/E1A complex that may both stabilise the E1A-pRb interaction (Barbeau,
Charbonneau et al. 1994) and promote acetylation of pRb at the C-terminus (Chan,
Krstic-Demonacos et al. 2001). This activity is important for E1A-induced cell cycle
progression and transformation (Egan et al, 1989; Whyte et al, 1989). CtBP, a putative
HDAC recruitment protein, binds to a region in the C-terminus (Goodman and Smolik
2000). Other known cellular binding partners include cyclin A, p400, CDK2, BS69, TBP
and various components of the TFIID complex.
1.9.3.2 E1B
E1A can induce apoptosis through the stabilisation of the p53 tumour suppressor protein
during oncogenic transformation (White, Sabbatini et al. 1992; Lowe and Ruley 1993;
White, Chiou et al. 1994), additional factor(s) are required to abrogate p53 and prevent
the induction of apoptosis. In adenovirus, this activity is performed by E1B. The ability
of adenovirus to segregate pRb and p53-abrogation activities between two different viral
oncoproteins is in contrast to the combined functional activity of LT.
Moreover,
multiple proteins are encoded by adenovirus to inhibit p53-dependent apoptosis; E1B55kDa and E4orf directly bind to and inactivate p53 (Yew, Liu et al. 1994; Nevels,
Rubenwolf et al. 1997), whereas E1B-19kDa blocks apoptosis by mimicking the antiapoptotic activity of Bcl2 (Rao, Debbas et al. 1992).
1.9.4 HPV Type 16
HPV type 16 is a member of the small double-strand DNA tumour virus family that
specifically infects squamous epithelial cells. The lifecycle of HPVs are linked to the
differentiation program of the host epithelial cells since HPVs infect undifferentiated,
basal keratinocytes, but most of the viral lifecycle occurs in the differentiated upper
epithelial strata where virus particles are shed. Papilloma viruses can be divided into
mucosal or cutaneous growth tropism groups and further subdivided in respect to their
56
propensity for malignant progression, namely high or low risk. HPV type 16 represents
the most prevalent mucosal high risk HPV type and two proteins encoded by HPV
function in an analogous manner to both LT and E1A and E1B, namely E6 and E7. E6
and E7 are stably expressed in HPV-positive cervical cancers and cancer-derived cell
lines (Schwarz, Freese et al. 1985), and they are essential to maintain the transformed
state of HPV-positive cells (Alvarez-Salas, Cullinan et al. 1998).
1.9.4.1 E7
E7 has been identified in approximately 90% of all human cervical cancers (zur Hausen
2001) and 20% of oral cancers (Gillison, Koch et al. 2000). E7 is a small multifunctional
protein of 98 aa encoded by the early region of HPV and is responsible for the ability of
this virus in overcoming G1 phase arrest induced by loss of cell adhesion, growth factor
withdrawl, DNA damage and differentiation signals. Similar to LT and E1A, E7
possesses the canonical LXCXE motif of Rb family binding in CR2 that facilitates the
targeted binding of hypophosphorylated pRb (Gage, Meyers et al. 1990) and other
members of the Rb family (Dyson and Harlow 1992; Davies, Hicks et al. 1993). In
addition to the LXCXE motif, E7 also shares sequence homology with LT and E1A in a
small region of CR1. E7 exhibits a high turnover rate of approximately 2 hours (hrs),
mediated by ubiquitin-mediated proteolysis.
A number of additional E7-interacting proteins have been described; for example, E7
can bind to two CDKIs, namely p21CIP1/WAF1/Sdi1and p27kip1. In addition, it has been
suggested that p21CIP1/WAF1/Sdi1 inactivation is critical for the ability of E7 to promote
viral DNA replication during keratinocyte differentiation, in addition to overriding the
cytostatic effect of TNF- in these cells. E7 has also been shown to associate indirectly
with cell cycle regulators such as cyclin A, cyclin E and CDK2 via p107 to promote
their aberrant expression and activity. There is also evidence that E7 can inhibit p53
transcriptional activity (Massimi and Banks 2000); in this model, Caseine kinase II
(CKII) activity may be required to phosphorylate E7 and stimulate the ability of E7 to
complex with TBP and form a tripartite complex with p53. This activity is similar to the
tripartite complex proposed for E1A, p53 and TBP.
57
1.9.4.2 E6
Like adenovirus, the activity of a second viral oncoprotein is required to directly bind to
and inactivate p53 to inhibit the induction of apoptosis. In HPV-16, E6, a small protein
of 151 aa mediates this activity (Scheffner, Munger et al. 1992). Unlike LT and E1B
however, E6 destabilises p53 by association with the ubiquitin ligase E6AP to promote
p53 degradation by the ubiquitination pathway (Huibregtse, Scheffner et al. 1991;
Scheffner, Huibregtse et al. 1993; Rapp and Chen 1998). This activity impairs the ability
of p53 to induce apoptosis or growth arrest; for example, the majority of human cervical
cancers exhibit wt p53 activity, yet its activity is functionally neutralised by the activity
of E6 (Thomas, Pim et al. 1999). The oncogenic activity of E6 has also been
demonstrated by its ability to transform established MEFs and to confer resistance to
terminal differentiation of human keratinocytes (Mantovani and Banks 2001). Its ability
to transcriptionally activate the catalytic component of human telomerase (hTERT) in
some cell types (Gewin and Galloway 2001; Oh, Kyo et al. 2001; Veldman, Horikawa et
al. 2001) is important for it to function as an oncoprotein. In addition to inactivating p53,
E6 can impair the activity of many other cellular proteins; for example, E6 downregulates p21CIP1/WAF1/Sdi1 in many normal cell types (Burkhart, Alcorta et al. 1999) and
interacts with the pro-apoptotic Bak, TNFR-1, and DNA repair proteins MGMT and
XRCC1 (Mantovani and Banks 2001; Filippova, Song et al. 2002), amongst others. The
fact that both pro- and anti-apoptotic activities for E6 have been described is difficult to
reconcile but may be cell context-dependent; for example, in HDFs, E6 expression
inhibits oxidant-induced apoptosis within 24 hrs but sensitises cells to apoptosis after
prolonged incubation (Chen and Wang 2000).
1.10
SASP: SENESCENCE-ASSOCIATED SECRETORY PHENOTYPE
AND ROS: REACTIVE OXYGEN SPECIES
It has long been known within the field that the culture medium of senescent cells is
enriched with secreted proteins (Shelton, Chang et al. 1999; Krtolica and Campisi 2002).
The SASP concept was first proposed by the Campisi group, when they realized that
58
secreted factors from senescent fibroblasts promote the transformation of pre-malignant,
but not of normal, mammary epithelial cells. This initial observation of SASP indicated
that senescence might not simply be a tumour suppressor mechanism, but rather a
double-edged sword within the tumour microenvironment. What remained unclear,
however, were the functional effects of SASP on the senescence phenotype itself. A
series of recent papers (Acosta, O'Loghlen et al. 2008; Coppe, Patil et al. 2008;
Kuilman, Michaloglou et al. 2008; Wajapeyee, Serra et al. 2008; Augert, Payre et al.
2009), have added various new members involved in SASP and notably IL-6 and IL-8
which are also up-regulated upon senescence in the HMF3A system described within
this thesis, and collectively reinforced the idea that senescence is both regulated by and
regulates the extracellular environment. Senescence bypass screening is a powerful tool
to identify new components of the senescence machinery. Some of these factors might
be potential tumour suppressors, whereas others could be 'context-dependent' tumour
suppressors or even oncogenes.
Senescence is clearly more complex than CDKI-mediated growth arrest or extrinsic
secretion signalling. Senescent cells express hundreds of genes differentially (Shelton,
Chang et al. 1999), prominent among these being pro-inflammatory secretory genes
(Coppe, Patil et al. 2008) and marker genes for a retrograde response induced by
mitochondrial dysfunction (Passos, Saretzki et al. 2007). Recent studies showed that
activated chemokine receptor CXCR2 (Acosta, O'Loghlen et al. 2008), insulin-like
growth factor binding protein 7 (Wajapeyee, Serra et al. 2008), IL6 receptor (Kuilman,
Michaloglou et al. 2008) or down-regulation of the transcriptional repressor HES1
(Sang, Coller et al. 2008) may be required for the establishment and/or maintenance of
the senescent phenotype in various cell types. A signature pro-inflammatory secretory
phenotype takes 7–10 days to develop under DDR (Coppe, Patil et al. 2008; Rodier,
Coppe et al. 2009). Together, these data suggest that senescence develops quite slowly
from an initiation stage (e.g. DDR-mediated cell cycle arrest) towards fully irreversible,
phenotypically complete senescence. It is the intermediary steps that define the
establishment of senescence, which are largely unknown with respect to kinetics and
governing mechanisms (Passos, Nelson et al. 2010).
59
Reactive oxygen species (ROS) are likely to be involved in establishment and
stabilization of senescence. Elevated ROS levels are associated with both replicative
(telomere-dependent) and stress- or oncogene-induced senescence (Saretzki 2003;
Ramsey and Sharpless 2006; Passos, Saretzki et al. 2007; Lu and Finkel 2008). ROS
also accelerate telomere shortening and can damage DNA directly and thus induce DDR
and senescence (Chen, Jackson et al. 1995; Lu and Finkel 2008; Rai, Phadnis et al.
2008). Conversely, activation of the major downstream effectors of the DDR/senescence
checkpoint can induce ROS production (Polyak, Xia et al. 1997; Macip, Igarashi et al.
2002; Macip, Igarashi et al. 2003).
Recently, a novel mechanism has been described for senescence; the existence of
a dynamic feedback loop that is triggered by a DNA damage response (DDR) and,
which after a delay of several days, locks the cell into an actively maintained state of
‗deep' cellular senescence. The essential feature of the loop is that long-term activation
of p21CIP1/WAF1/Sdi1 induces mitochondrial dysfunction and production of reactive oxygen
species (ROS) through serial signalling including GADD45-MAPK14 (p38MAPK)GRB2-TGFBR2-TGFβ. These ROS in turn replenish short-lived DNA damage foci and
maintain an ongoing DDR. This loop was shown to be both necessary and sufficient for
the stability of growth arrest during the establishment of the senescent phenotype.
1.11
NF-B PATHWAY
1.11.1 Introduction
NF-κB was first discovered in the lab of Nobel Prize laureate David Baltimore via its
interaction with an 11-base pair sequence in the immunoglobulin Kappa light-chain
enhancer in B cells and plasma cells but not pre B-cells (Sen and Baltimore 1986).
Later, it was demonstrated that NF-B DNA binding activity was induced by a variety
of extrinsic factors, and that this activation is independent from de-novo protein
60
synthesis. NF-B has been detected in most cell types, and specific NF-B binding sites
have been identified in promoters and enhancers of a high number of inducible genes.
NF-κB proteins comprise a family of structurally-related eukaryotic transcription factors
that are involved in the control of a large number of normal cellular and organismal
processes, such as immune and inflammatory responses, stress and injury. Some
examples are the induction of IL-2, TAP1 and MHC molecules and involvement in
many aspects of the inflammatory response, such as induction of IL-1 (alpha and beta),
TNF-alpha and leukocyte adhesion molecules (E-selectin, VCAM-1 and ICAM-1). NFκB is also involved in many aspects of cell growth, differentiation, proliferation and
apoptosis via the induction of certain growth and transcription factors (e.g. c-myc, ras
and p53). In addition, these transcription factors are persistently active in a number of
disease
states,
including
cancer,
arthritis,
chronic
inflammation,
asthma,
neurodegenerative diseases, and heart disease.
1.11.2 NF-κB family
There are five proteins in the mammalian NF-κB family (Nabel and Verma 1993): RelA,
RelB, c-Rel, NFKB1 and NFKB2. All NF-κB family members share structural
homology with the retroviral oncoprotein v-Rel, resulting in their classification as NFκB / Rel proteins (Gilmore 2006). RelA, RelB, and c-Rel all have a transactivation
domain in their C-terminus. In contrast, the NFKB1 and NFKB2 proteins are
synthesized as large precursors, p105, and p100, which undergo processing to generate
the mature NF-κB subunits, p50 and p52, respectively. The processing of p105 and p100
is mediated by the ubiquitin/proteasome pathway and involves selective degradation of
their C-terminal region containing ankyrin repeats. Whereas the generation of p52 from
p100 is a tightly-regulated process, p50 is produced by constitutive processing of p105
(Karin and Ben-Neriah 2000; Senftleben, Cao et al. 2001).
61
1.11.3 Activation
Part of NF-κB's importance in regulating cellular responses is that it belongs to the
category of "rapid-acting" primary transcription factors, i.e., transcription factors that are
present in cells in an inactive state and do not require new protein synthesis to be
activated (other members of this family include transcription factors such as cJun, STATs, and nuclear hormone receptors). This allows NF-κB to act as a "first
responder" to harmful cellular stimuli. Stimulation of a wide variety of cellsurface receptors, such as RANK, TNFR, leads directly to NF-κB activation and fairly
rapid changes in gene expression (Gilmore 2006)
NF-κB can be induced by stimuli such as pro-inflammatory cytokines and bacterial
toxins (e.g. LPS, exotoxin B) and a number of viruses/viral products (e.g. HIV-1,
HTLV-I, HBV, EBV, Herpes simplex) as well as pro-apoptotic and necrotic stimuli
(oxygen free radicals, UV light, gamma-irradiation). Many bacterial products, as an
example, can activate NF-κB. The identification of Toll-like receptors (TLRs) as
specific pattern recognition molecules and the finding that stimulation of TLRs leads to
activation of NF-κB improved our understanding of how different pathogens activate
NF-κB. For example, studies have identified TLR4 as the receptor for the LPS
component of Gram-Negative bacteria (Doyle and O'Neill 2006). TLRs are key
regulators of both innate and adaptive immune responses (Hayden, West et al. 2006).
Unlike RelA, RelB, and c-Rel, the p50 and p52 NF-κB subunits do not contain
transactivation domains in their C terminal halves. Nevertheless, the p50 and p52 NF-κB
members play critical roles in modulating the specificity of NF-κB function. Although
homodimers of p50 and p52 are, in general, repressors of NF-κB site transcription; both
p50 and p52 participate in target gene transactivation by forming heterodimers with
RelA, RelB, or c-Rel (Li and Verma 2002). In addition, p50 and p52 homodimers also
bind to the nuclear protein Bcl-3, and such complexes can function as transcriptional
activators (Franzoso, Bours et al. 1992; Bours, Franzoso et al. 1993; Fujita, Nolan et al.
1993).
62
1.11.4 Inhibition
In unstimulated cells, the NF-κB dimers are sequestered in the cytoplasm by a family of
inhibitors, called IκBs (Inhibitor of κB), which are proteins that contain multiple copies
of ankyrin. By virtue of their ankyrin repeat domains, the IκB proteins mask the nuclear
localization signals (NLS) of NF-κB proteins and keep them sequestered in an inactive
state in the cytoplasm (Jacobs and Harrison 1998).
1.11.4.1
The IκB family
To date seven IκB s have been identified: IκBα, IκBβ, IκBγ, IκBε and Bcl-3 but the beststudied and major IκB protein is IκBα. Due to the presence of ankyrin repeats in their Cterminal halves, p105 and p100 also function as IκB proteins. IκBγ is unique in that it is
synthesized from the NFKB1 gene using an internal promoter, thereby resulting in a
protein that is identical to the C-terminal half of p105 (Inoue, Kerr et al. 1992). The cterminal half of p100, that is often referred to as IκBδ, also functions as an inhibitor
(Dobrzanski, Ryseck et al. 1995; Basak, Kim et al. 2007).
1.11.4.2
IκB kinase: IKK
Activation of the NF-κB is initiated by the signal-induced degradation of IκB proteins:
signals that induce NF-κB activity cause the phosphorylation of IκBs, their dissociation
and subsequent degradation, allowing NF-κB proteins to enter the nucleus and induce
gene expression.
This occurs primarily via activation of a kinase called the IκB kinase (IKK). IKK is
composed of a heterodimer of the catalytic IKK-alpha and IKK-beta subunits and a
"master" regulatory protein termed NEMO (NF-κB essential modulator) or IKK-gamma
(Figure 1.6). When activated by signals, usually coming from the outside of the cell, the
IκB kinase phosphorylates two serine residues located in an IκB regulatory domain
(serines 32 and 36 in human IκBα) leading to the ubiquitinylation of the IκB inhibitor
molecules and their degradation by the proteasome.
63
Figure 1.6: NF-κB: The canonical pathway
The binding of liguand to a receptor leads to the recruitment and activation of an IKK complex comprising
IKK alpha and/or IKK beta catalytic subunits and two molecules of NEMO. The IKK complex then
phosphorylates IkB leading to its degradation by the proteasome. NFkB then translocates to the nucleus to
activate target genes regulated by kB sites.
64
With the degradation of the IκB inhibitor, the NF-κB complex is then free to enter the
nucleus where it can 'turn on' the expression of specific genes that have DNA-binding
sites for NF-κB. The activation of these genes by NF-κB then leads to the given
physiological response, for example, an inflammatory or immune response, a cell
survival response, or cellular proliferation.
IKK-alpha knockout mice die shortly after birth and exhibit developmental
abnormalities such as shortened and truncated limbs, ears, heads and snouts due to a
defect of differentiation of skin epidermal cells (keratinocytes). In general, IKK-alpha
seems to be involved in skeletal development. Interestingly, IL-1 and TNF-alpha still
can activate NF-κB in cells from IKK-alpha -/- mice.
IKK-beta knockout mice-embryos die from excessive loss of hepatocytes due to
apoptosis. Apoptosis appears to be induced by TNF-alpha since IKK-beta and TNFR1
double knockout mice are not affected by hepatocyte apoptosis and embryonic death.
Additionally, fibroblasts from IKK-beta -/- mice undergo apoptosis in response to TNFalpha, presumably due to a missing "survival" signal usually provided by NF-κB
activation (May and Gosh, 1999).
1.11.5 Canonical NF-B pathway
There are two signalling pathways leading to the activation of NF-B known as
the canonical pathway (or classical) and the non-canonical pathway (or alternative
pathway) (Karin 1999; Gilmore 2006; Scheidereit 2006; Tergaonkar 2006). The
common regulatory step in both of these cascades is activation of an IκB kinase (IKK)
complex consisting of catalytic kinase subunits (IKKa and/or IKKb) and the regulatory
non-enzymatic scaffold protein NEMO (NF-kappa B essential modulator also known as
IKKg) (Figure 1.6). Activation of NF-B dimers is due to IKK-mediated
phosphorylation-induced proteasomal degradation of the IκB inhibitor enabling the
active NF-B transcription factor subunits to translocate to the nucleus and induce target
65
gene expression. NF-B activation leads to the expression of the IκBa gene, which
consequently sequesters NF-B subunits and terminates transcriptional activity unless a
persistent activation signal is present.
In the canonical signalling pathway, binding of ligand such as IL-1 or TNFalpha to a cell
surface receptor such as a member of the Toll-like receptor super-family leads to the
recruitment of adaptors (such as TRAF) to the cytoplasmic domain of the receptor
(Figure 1.6). These adaptors, in turn, recruit the IKK complex comprising IKK alpha
and/or IKK beta catalytic subunits and two molecules of NEMO. This leads to the
phosphorylation and degradation of the IκB inhibitor. The canonical pathway activates
NF-B dimers comprising of RelA, c-Rel, RelB and p50.
1.11.6 Non-canonical pathway
In this alternative NF-B activation pathway, activation of NIK (NF-κB inducing
kinase) upon receptor ligation leads to the phosphorylation and subsequent proteasomal
processing of the NFKB2 precursor protein p100 into mature p52 subunit (Figure 1.7).
Then p52 dimerizes with RelB to appear as a nuclear RelB/p52 DNA binding activity
and regulate a distinct class of genes (Bonizzi, Bebien et al. 2004). In contrast to the
canonical signalling that relies upon NEMO-IKK mediated degradation of IκBα, -β, -ε,
the non-canonical signalling critically depends on NIK mediated processing of p100 into
p52. This pathway utilizes an IKK complex that comprises two IKKa subunits, but not
NEMO. Given their distinct regulations, these two pathways were thought to be
independent of each other. However, recent analyses revealed that synthesis of the
constituents of the non-canonical pathway, RelB and p52, is controlled by the canonical
IKK-IκB-RelA/p50 signalling (Basak, Shih et al. 2008). This suggests that an integrated
NF-κB system network underlies activation of both RelA and RelB containing dimer
and that a malfunctioning canonical pathway will lead to an aberrant cellular response
also through the non-canonical pathway.
66
Figure 1.7: NF-κB: The non canonical pathway
Receptor binding leads to the activation of NIK, which phosphorylates and activates an IKK alpha
complex that in turn phosphorylates the IκB domain of p100 leading to the liberation of p52/RelB. This
heterodimer subsequently translocates to the nucleus to activate target genes regulated by κB sites.
67
1.11.7 NF-B and cancer
In many cancer cells (including breast cancer, colon cancer, prostate cancer, lymphoid
cancers, and probably many others; see Diseases link) NF-B is constitutively active and
located in the nucleus. In some cancers, this is due to chronic stimulation of the IKK
pathway, while in other cases (such as some Hodgkin's and diffuse large B-cell
lymphoma cells) the gene encoding IκB can be mutated and defective. Moreover,
several human lymphoid cancer cells have mutations or amplifications of genes
encoding Rel / NF-B transcription factors (REL in human B-cell lymphoma) and many
multiple myelomas have mutations in genes encoding NF-B signaling regulatory
proteins that lead to constitutive activation of NF-B. It is thought that continuous
nuclear Rel / NF-B activity protects cancer cells from apoptosis and in some cases
stimulates their growth. Therefore, many current anti-tumour therapies seek to block
NF-B activity as a means to inhibit tumour growth or to sensitize the tumour cells to
more conventional therapies, such as chemotherapy.
1.11.8 NF-B, senescence and ageing
The NF-B family of ubiquitously transcription factors is widely known as key
regulators of inflammatory and immune response. However, more recently they have
been shown to function as regulators of diverse cellular processes such as cell
proliferation and differentiation and the response to stresses such as oxidative, physical
and chemical stress. Activation of NF-B also blocks apoptosis and promotes cell
survival.
A previous study in our lab suggested that the loss of proliferative potential in the
HMF3A conditionally immortal fibroblasts may involve the activation of the NF-B
pathway (Hardy, Mansfield et al. 2005). NF-B has also been shown to be associated
68
with growth arrest in the study of Penzo (Penzo, Massa et al. 2009) who has shown that
acute activation of NF-B in murine embryo fibroblasts results in growth arrest. The
growth arrest was associated with repression of 20 genes essential for cell cycle
progression that are known targets of either E2F or FOXM1.
In addition, Adler (Adler, Sinha et al. 2007) using a systematic bioinformatic approach
to identify combinatorial cis-regulatory motifs showed that NF-B activity controlled
cell cycle exit and was continually required to enforce many features of ageing in a
tissue specific manner. Moreover activation of NF-B with age is consistent with
elevated levels inflammatory markers and a pro-inflammatory phenotype associated with
many age related diseases. Factors that mediate NF-B and inflammation include the
insulin/IGF pathway, SIRT1, FOXO, PDC-1 and PPAR (Salminen, Ojala et al. 2008).
Expression of relA was found to be lower in senescent cells (Helenius, Hanninen et al.
1996) whereas c-Rel was elevated (Bernard, Gosselin et al. 2004). Kriete (Kriete, Mayo
et al. 2008) showed that there was a constitutive activation of NF-B in older human
subjects compared to young donors.
1.12
MICRO-RNAS
1.12.1 Introduction
Micro-RNAs are a class of post-transcriptional regulators (Kusenda, Mraz et al. 2006;
Vasudevan, Russell et al. 2007; Place, Li et al. 2008). They are short ~22 nucleotide
RNA sequences that bind to fully or partially complementary sequences in the 3‘
UTR of multiple target mRNAs, usually resulting in their silencing (Bartel 2004).
Micro-rnas have been predicted to target ~60% of all genes (Friedman, de Jong et al.
2007), are abundantly present in all human cells (Lim, Lau et al. 2003) and are able to
repress hundreds of targets (Brennecke, Stark et al. 2005).
69
Micro-rnas were first discovered in 1993 by Victor Ambros, Rosalind Lee and Rhonda
Feinbaum during a study into development in the nematode C. elegans regarding the
gene lin-14 (Lee, Feinbaum et al. 1993). This screen led to the discovery that the gene
lin-14 was able to be regulated by a short RNA product from lin-4 itself, a gene that
transcribed a 61 nucleotide precursor that matured to a 22 nucleotide mature RNA which
contained sequences partially complementary to multiple sequences in the 3‘ UTR of the
lin-14 mRNA. This complementarity was sufficient and necessary to inhibit the
translation of lin-14 mRNA. Since then, over 10,000 miRNAs have been discovered in
all studied multicellular eukaryotes including mammals, fungi and plants. More than 700
miRNAs have so far been identified in humans (www.miRbase.com) and over 800 more
are predicted to exist (Bentwich, Avniel et al. 2005).
Due to their abundant presence and far-reaching potential, miRNAs have all sorts of
functions in physiology, from cell differentiation, proliferation, apoptosis (Brennecke,
Hipfner et al. 2003) to the endocrine system (Poy, Eliasson et al. 2004), haematopoiesis
(Chen, Li et al. 2004), fat metabolism (Wilfred, Wang et al. 2007). They display
different expression profiles from tissue to tissue (Lagos-Quintana, Rauhut et al. 2002),
reflecting the diversity in cellular phenotypes and as such suggest a role in tissue
differentiation and maintenance.
1.12.2 MiRNA and siRNA
Micro-rna are similar to, but distinct from, another type of short RNA, known as small
interfering RNA (siRNA). Although miRNA and siRNA both have gene regulation
functions, there are subtle differences. MiRNA may be slightly shorter than siRNA
(which has 20 to 25 nucleotides). MiRNA is single-stranded, while siRNA is formed
from two complementary strands. The two kinds of RNA are encoded slightly
differently. siRNA are usually synthesised in vitro and introduced by transfection but
can also be generated from shRNA or from miRNAs. The mechanism by which they
regulate genes is also slightly different.
70
1.12.3 Biogenesis
Micro-RNA genes reside in regions of the genome as distinct transcriptional units as
well as in clusters of polycistronic units - carrying the information of several micro-rnas
(Lagos-Quintana, Rauhut et al. 2001; Lau, Lim et al. 2001; Reinhart, Weinstein et al.
2002). Studies suggest that approximately half of known micro-RNA reside in nonprotein coding RNAs (intron and exon) or within the intron of protein coding genes
(Rodriguez, Griffiths-Jones et al. 2004), generally within the 3‘UTR. Micro-rna
(miRNA) genes are generally transcribed by RNA Polymerase II (Pol II) in the nucleus
to form large primary micro-rna transcripts (pri-miRNA) (Figure 1.8), which are capped
and polyadenylated (Kim 2005). These pri-miRNA transcripts are then processed into
micro-rna precursor (pre-miRNA) by the microprocessor complex Drosha–DGCR8
(Pasha) in the nucleus. The resulting precursor hairpin, the pre-miRNA, is exported from
the nucleus by Exportin-5–Ran-GTP. In the cytoplasm, the RNase Dicer in complex
with the double-stranded RNA-binding protein TRBP cleaves the pre-miRNA hairpin to
its mature length. The functional strand of the mature miRNA is loaded together with
Argonaute (Ago2) proteins into the RNA-induced silencing complex (RISC), where it
guides RISC to silence target mRNAs through mRNA cleavage, translational repression
or deadenylation, whereas the passenger strand (black) is degraded.
1.12.4 Mechanism of MiRNA regulation
Once incorporated into a RISC, the mature micro-rna binds to the mRNA target to
negatively regulate gene expression in one of two ways that depend on the degree of
complementarity between the miRNA and its target:

miRNAs that bind to their mRNA targets with perfect (or nearly perfect)
complementarity induce target-mRNA direct cleavage and destruction of the
mRNA (Rhoades, Reinhart et al. 2002; Chen and Meister 2005) most usually in
plants. miRNAs using this mechanism bind to miRNA complementary sites that
are generally found in the coding sequence or ORF of the mRNA target.
71
Figure 1.8: Micro-RNAs biogenesis
MicroRNA (miRNA) genes are generally transcribed by RNA Polymerase II (Pol II) in the nucleus to
form large primary microRNA transcripts (pri-miRNA, which are capped and polyadenylated. These primiRNA transcripts are then processed into microRNA precursor (pre-miRNA) by the microprocessor
complex Drosha–DGCR8 (Pasha) in the nucleus. The resulting precursor hairpin, the pre-miRNA, is
exported from the nucleus by Exportin-5–Ran-GTP. In the cytoplasm, the RNase Dicer in complex with
the double-stranded RNA-binding protein TRBP cleaves the pre-miRNA hairpin to its mature length. The
functional strand of the mature miRNA is loaded together with Argonaute (Ago2) proteins into the RNAinduced silencing complex (RISC), where it guides RISC to silence target mRNAs through mRNA
cleavage, translational repression or deadenylation, whereas the passenger strand (black) is degraded.
72

In contrast, nearly all animal miRNAs studied so far are not usually exactly
complementary to their mRNA targets, and seem to inhibit protein synthesis
while retaining the stability of the mRNA target (Ambros 2004). miRNAs that
bind to mRNA targets with imperfect complementarity block target gene
expression at the level of protein translation. Recent evidence indicates that
miRNAs might also affect mRNA stability. Complementary sites for miRNAs
using this mechanism are generally found in the 3' untranslated regions (3'
UTRs) of the target mRNA genes.
It has been suggested that transcripts may be regulated by multiple miRNAs and that an
individual miRNA may target numerous transcripts if their sequences have similarities.
It all depends on the seed sequence which is formed by seven or eight nucleotides of the
mature miRNA, starting from the first or second nucleotide, and is most crucial for
interaction between the miRNA and its target.
1.12.5 Micro-RNAs and cancer
The relevance of miRNAs to cancer was suggested by changes in their expression
patterns (Iorio, Ferracin et al. 2005; Volinia, Calin et al. 2006) and recurrent
amplification and deletion of miRNA genes in tumours (Akao, Nakagawa et al. 2006;
Dews, Homayouni et al. 2006).
Several miRNAs have emerged as candidate components of oncogene and tumoursuppressor networks. The miR-17-92 cluster (He, Thomson et al. 2005; O'Donnell,
Wentzel et al. 2005; Dews, Homayouni et al. 2006), miR-372/373 (Voorhoeve, le Sage
et al. 2006) and miR-155/BIC (Tam and Dahlberg 2006) have been implicated as protooncogenes in B‑cell lymphomas and testicular cancers. On the other hand, miR-15-16 is
frequently deleted in patients with chronic lymphocytic leukaemia (CLL) (Calin,
Dumitru et al. 2002; Mraz, Pospisilova et al. 2009). Expression studies and functional
studies have also revealed the potential tumour-suppressive roles of let-7 in various
73
cancers (Johnson, Grosshans et al. 2005; Mayr, Hemann et al. 2007), possibly owing to
its ability to repress key oncogenic components, including Ras and HMGA2.
A study of mice altered to produce excess c-myc — a protein implicated in several
cancers — shows that miRNA has an effect on the development of cancer. Mice
engineered to produce a surplus of types of miRNA namely the cluster mir-17–92 found
in lymphoma cells developed the disease within 50 days and died two weeks later. In
contrast, mice without the surplus miRNA lived over 100 days (Cui, Li et al. 2007).
Another study found that two types of miRNA (miR 17-5p and miR-20b) inhibit the
E2F1 protein, which regulates cell proliferation. miRNA appears to bind to messenger
RNA before it can be translated to proteins that switch genes on and off (O'Donnell,
Wentzel et al. 2005).
Consistent with this, the suppression of key components of the miRNA biogenesis
machinery in cancer cells has been reported to promote transformation both in vitro and
in vivo (Kumar, Lu et al. 2007). The true extent to which the disruption of miRNA
pathways has a role in tumourigenesis remains to be determined. However, early
indications are that this family of genes is intimately integrated into the regulatory
processes that are normally disrupted during transformation. Moreover, the placement of
several miRNAs into known oncogenic and tumour-suppressor networks is beginning to
solve longstanding mysteries of how the circuitry of these pathways is wired.
1.12.6 Micro-RNA and senescence
Several studies have started linking micro-RNA regulation and cellular senescence but
the exact mechanisms of this relation remains to be specified.
The ability of miRNAs to regulate a variety of target genes allows them to induce
changes in multiple pathways and processes such as development, apoptosis,
proliferation and differentiation. MiRNAs could therefore facilitate the complex cellular
changes required to establish the senescent phenotype. Identification of the mRNA
74
sequences that miRNAs regulate is mainly derived using bioinformatics techniques. The
mirBase sequence database is the main repository for miRNA sequence and target
information and contains 695 human miRNA sequences, each with the potential to
regulate on average 1000 gene targets. It is this large number of potential targets across
different biological pathways that could give miRNAs the power to potentially induce
complex cell phenotypes, like senescence.
Several studies highlight a number of senescence-associated micro-rnas such as Let-7f,
miR-499, miR-371 (Wagner, Horn et al. 2008), miR-372, miR-373 (Voorhoeve, le Sage
et al. 2006), miR-34a (He, He et al. 2007; Tazawa, Tsuchiya et al. 2007), miR-34b and
miR-34c (Kumamoto, Spillare et al. 2008), miR-20b (Poliseno, Pitto et al. 2008) In
addition, tumour-suppressive miR-34a expression induced senescence-like growth arrest
through modulation of the E2F pathway in human colon cancer cells (Tazawa, Tsuchiya
et al. 2007).
1.13
MODEL OF STUDY: HMF3A CELLS
One of the main stumbling blocks in studying the molecular pathways that underlie the
finite proliferative life span has been the absence of suitable model systems for study
because of the asynchrony of this process in heterogeneous cell populations that are
typically used for serial sub-cultivation.
Studies with human cells are further
complicated by the genetic, epigenetic and proliferative variation that can exist between
different donors as well as phenotypic differences between cells within the cultures. To
simplify this process many investigators study oncogene-induced senescence (OIS) with
the expression of activated oncogenes such as RASV12, RAF, BRAF or E600 since it
occurs prematurely without telomere attrition and can be induced acutely in a variety of
cell types (Serrano, Lin et al. 1997; Michaloglou, Vredeveld et al. 2005; Collado and
Serrano 2010).
A different approach was taken by making use of the finding that reconstitution of
telomerase activity by introduction of the catalytic subunit of human telomerase
(hTERT) alone was incapable of immortalising all human somatic cells (Bodnar,
75
Ouellette et al. 1998; Vaziri, Ni et al. 1998), but inactivation of the p16-pRb and p53p21 pathways were required in addition (Counter, Meyerson et al. 1998; Kiyono, Foster
et al. 1998). It was found that expressing LT, a viral oncoprotein that binds and inhibits
the activity of several proteins, including p53 and Rb (Ali and DeCaprio 2001), together
with expressing hTERT, can immortalize human primary mammary fibroblasts by
preventing cellular senescence (O'Hare, Bond et al. 2001) (Figure 1.9).
This observation permitted the use a thermolabile mutant (U19tsA58) of LT antigen to
develop a line of conditionally immortalised human mammary fibroblasts (HMF3A) that
remain stringently temperature sensitive and show no sign of transformation in >300
population doublings (Figure 1.9). These cells are immortal if grown at 34C but
undergo an irreversible growth arrest within 5 days upon shift up to 38C when the
thermolabile T antigen is inactivated (O'Hare, Bond et al. 2001). When these cells cease
dividing, SA-β-Gal activity is induced and the growth-arrested cells have features and
express genes in common with senescent cells (Hardy, Mansfield et al. 2005). Since
these cells growth arrest in a synchronous manner they are potentially an excellent
starting point for dissecting the pathways that underlie cellular senescence and act
downstream of p16-pRb and p53-p21.
For these reasons, the conditionally immortalised HMF3A system represented a
potential system with which to dissect telomere-independent cellular senescence
pathways by determining target genes ability to complement the growth of these cells
under non permissive conditions.
1.13.1 Reconstitution of WT LT activity in the HMF3A system alone
The conditionally immortalised phenotype of the HMF3A system is critically dependent
upon the activity of U19tsA58 LT (Hardy et al, 2005; O'Hare et al, 2001). Dr. Louise
Mansfield has shown in her thesis that reconstitution of wt LT activity into these cells by
amphotropic retroviral infection was sufficient to overcome conditional senescence in a
stable manner.
76
Figure 1.9: Engineering of the CL3EcoR cells
CL3EcoR cells were engineered by using a thermolabile mutant (U19tsA58) of LT antigen along
with expressing the human catalytic subunit of human telomerase (hTERT) to develop a line of
conditionally immortalized human mammary fibroblasts (HMF3A) . In a second time, the
HMF3A was refined by expressing a murine ecotropic receptor is a stable manner. Finally, after
introduction of the receptor, the cells have also been cloned to produce a consistent and
homogenous population.
77
1.13.2 Refinement of the HMF3A system by introduction of the murine
ecotropic receptor
The HMF3A system was initially refined by Dr. Louise Mansfield by engineering the
HMF3A cells to express a murine ecotropic receptor so that they become infectable with
ecotropic retroviruses. This step has increased the cell transduction efficiency and the
safety of the manipulations due to the fact that ecotropic viruses, unlike amphotropic
viruses, cannot infect human cells.
1.14
ABROGATION OF THE P53 PATHWAY
Louise Mansfield also showed that abrogation of the p53 pathway into the mixed
population of HMF3AEcoR by ecotropic retroviral delivery of either p21CIP1/WAF1/Sdi1
shRNA, p53 GSE or p53 shRNA was sufficient to bypass the conditional growth defect.
1.15
ABROGATION OF THE PRB PATHWAY
Similarly to p53, Dr Louise Mansfield tried to target the pRb pathway for inactivation.
Louise Mansfield showed that abrogation of the pRb pathway into the mixed population
of HMF3AEcoR by ecotropic retroviral expression of E1A or E7 was sufficient to
bypass the conditional growth defect.
However, the induction could not be attributed to pRb alone as both expression
constructs used, namely E7 and E1A, could have used their multifunctional activity to
abrogate other pathways as well as the pRb pathway. As a consequence, Dr. Mansfield
tried to inactivate the pRb pathway using other reagents. The results of this investigation
are presented here briefly for further understanding of the research strategy detailed in
the thesis.
78
1.15.1 Inactivation of the INK4A Locus
In addition to E7 and E1A, and in order to investigate whether the abrogation of the Rb
pathway specifically was sufficient to bypass the growth arrest, Dr. Louise Mansfield
tried to inactivate the INK4A locus. Two splice variants are encoded by the INK4A
locus; the cyclin-dependent kinase inhibitor p16INK4a and p14ARF. Both of these splice
variants represent components of the pRb pathway: p16INK4a negatively regulates pRb
functional activity by inhibiting cyclin D-CDK4/6 complexes, whereas p14 ARF function
downstreams of pRb to negatively regulate pRb effector signalling in an E2F-dependent
process. Furthermore, p14ARF acts as a link between the pRb and p53 pathways as it
stabilises p53 by binding to the Mdm2 protein.
Therefore, it was anticipated that abrogation of either, or even both, of these components
would functionally inactivate the pRb pathway. Dr Louise Mansfield performed these
experiments in the HMF3AEcoR cells.
1.15.1.1
Knockdown of p14ARF by ShRNA
Knockdown of p14ARF was performed by Dr. Louise in the HMF3A system with a
construct found to silence p14ARF in HDF (Berns, Hijmans et al. 2004). The
complementation did not work in this case compared to the positive control, p53
shRNA. This indicated that p14ARF knockdown was insufficient to overcome the
HMF3A conditional growth arrest.
1.15.1.2
Knockdown of p16INK4a by ShRNA
In a similar manner, p16INK4a knockdown by shRNA was investigated by Dr. Louise
Mansfield with two constructs that were previously shown to work and to be insufficient
alone to bypass the induction of senescence in other cell types(Wei, Herbig et al. 2003;
Reynolds, Leake et al. 2004), but in the HMF3A, only a small reduction or no reduction
79
in p16INK4a protein levels could be observed by Western blot analysis. Consequently, the
level of p16INK4a knockdown was not considered significant.
p16INK4a knockdown was also proved insufficient to overcome conditional growth arrest
in BJ cells constitutively expressing hTERT and a temperature sensitive mutant of LT
(BJ-TERT-tsLT cells); (Berns, Hijmans et al. 2004).
1.15.1.3
Constitutive Expression of Bmi-1
Since RNAi could not effectively knock down p16INK4a at the protein level, an
alternative method to inactivate the INK4A locus was sought by Dr. Louise Masfield.
Bmi-1 is a transcriptional repressor of the PcG family that promotes stable, epigenetic
gene silencing though chromatin modifications mediated by histone methylation (van
der Lugt, Domen et al. 1994). Its constitutive expression leads to the inactivation of the
INK4A locus. Bmi-1 has been shown to be significantly down-regulated upon
replicative senescence in primary HDFs, but not in quiescent HDFs, whereas its overexpression was sufficient to extend the replicative lifespan of some HDF strains (Jacobs,
Kieboom et al. 1999; Itahana, Zou et al. 2003). A Bmi-1 retroviral expression construct,
pBabepuro-Bmi-1 was introduced into HMF3AEcoR cells by Dr. Louise Mansfield and
assessed for its ability to complement the conditional growth defect of these cells.Not
only did the complementation not work but Western blot analysis also revealed that
ectopic expression of Bmi-1 did not affect the steady-state levels of p16INK4a protein.
1.16
AIM OF THE THESIS
Due to the asynchronous nature of the growth arrest, senescence is a difficult
process to study in serially sub-cultivated primary human cells.
Therefore, a
conditionally immortalised human mammary fibroblast cell line was developed in the
JAT laboratory by retroviral transduction of early passage, adult interlobular mammary
fibroblasts with a temperature sensitive (ts), non-DNA-binding mutant SV40 LT,
80
namely U19tsA58, and hTERT (O'Hare, Bond et al. 2001). The HMF3A cells have been
modified by the stable expression of an ecotropic receptor allowing more efficient and
safe use of ecotropic virus supernatants in the cells. After introduction of the receptor,
the cells have also been cloned to produce a consistent and homogenous population.
This represented a good model in which to study cellular senescence.
The aims of the thesis were the following:
1. Validation of the HMF3AEcoR clone 3 or CL3EcoR
It was important to specify the senescence model and test the sensitivity of the
complementation assay in the mixed population of HMF3AEcoR cells and the clonal
model, CL3EcoR, to compare their sensitivity, consistency and representativity. Another
big objective was to optimise the complementation assay to be more standardised and
with a minimal background.
2. Validation of the complementation assay using p53 and pRb abrogation
alone
Here, I wanted to confirm that abrogation of either the p53 pathway alone or the pRb
alone were both sufficient to bypass the conditional growth arrest in both cell models.
3. To identify the changes in gene expression triggered by senescence by
expression profiling and validate NF-B involvement in senescence
Since these cells growth-arrest in a synchronous manner, Affymetrix expression
profiling was used to identify the genes differentially expressed specifically upon
senescence. I also wanted to validate some of the identified targets in vitro.Hardy et al
have shown previously, thanks to an in silico promoter analysis that cEBPbeta and NF-
B might be activated upon senescence (Hardy, Mansfield et al. 2005). I wanted to
validate that this activation was real in this model both by the expression profiling and in
vitro validation.
81
4. To look for new targets by using a shRNA functional screen using
complementation assay
Since a complementation assay was optimised and validated in these cells and had
permitted successfully to check the importance of several genes from the pRb or p53
pathway, the next step was to widen this principle to a gene library in order to identify
new senescence key effectors, not necessarily already described as such in the literature.
I wanted to apply a retroviral shRNA screen covering ~10,000 genes with the same cell
model by complementation assay.
5. To identify the changes in micro-rnas expression triggered by senescence
by expression profiling
At the beginning of the thesis, the literature was submerged by articles about the new
forefront micro-RNAs represented in gene expression modulation responsible of various
cellular mechanism and diseases. I consequently planned to profile micro-RNAs
expression upon senescence in the model thanks to MiRNA microarray technology (LC
Sciences). I also wanted to validate the identified targets in vitro.
6. Identify miRs targets and effect in vitro
I then wanted to see what effect these validated micro-rnas targets had on the
transcriptome by expression profiling and in silico analysis and whether these new data
would overlap with previous data obtained within this thesis.
82
2
MATERIAL AND METHODS
2.1 MAMMALIAN CELL CULTURE
2.1.1 Cell lines and Culture
HEK293T, Phoenix Eco and Phoenix Ampho were obtained from the ATCC. HMF3A
and HMF3AECoR cells were a proprietary cell line (O'Hare, Bond et al. 2001). BJ
primary cells were obtained from ATCC.
2.1.2 Cell media
Tissue culture media and cell culture reagents were purchased from Invitrogen. All
HMF3A and primary cells were maintained in Dulbecco‘s Modified Eagle Medium
(DMEM) supplemented with 2 millimolar (mM) glutamine, 100 units/ml penicillin, 100
g/ml streptomycin and 10% volume per volume (v/v) heat inactivated foetal calf serum
(FCS).
Primary cells, after infection with RAS, were maintained in the same medium but phenol
free and the FCS was replaced by heat inactivated charcoal stripped FBS (fetal bovine
serum).
2.1.3 Cell Culture Conditions
All cell lines were maintained in a 5% CO2 and 20% oxygen atmosphere.

amphotropic and  ecotropic cell lines were maintained at 37C. The HMF3A cell line
and HMF3AEcoR cells were maintained at 34C ±0.5C, a temperature at which the cells
83
proliferated continuously due to the functional activity of U19tsA58 LT.
HMF3A
temperature shift experiments were performed at 38C ±0.5C, a temperature at which
U19tsA58 LT was inactivated and the cells became senescent within a period of 5 days
(O‘Hare et al, 2001). Primary cells were maintained at 37C ±0.5C.
2.1.4 Sub-Culturing of Cells
Cells were grown until a sub-confluent state was reached (approximately 80%
confluence). Media was then removed and the monolayer of cells was washed twice
with PBS. The monolayer was detached using 1x trypsin-EDTA (1 ml/T75 cm2 flask)
for 5 mins at 34C and the trypsin-EDTA was inactivated by adding 10 ml of complete
media. Cells were then plated at a defined ratio (e.g. 1 in 8 of the total cells), or counted
using a haemocytometer and plated at the required density.
2.1.5 Preservation of Cells
Cells from a sub-confluent T75 cm2 flask were trypsinised, resuspended in complete
media and spun down at 1200 revolutions per minute (rpm) for 2 mins to remove any
traces of trypsin. Cells were resuspended in FCS supplemented with 10% dimethyl
sulphoxide (DMSO; BDH). 2x 1 ml aliquots were then transferred to cryotubes and
frozen at -70C wrapped in several layers of tissue for insulation.
Tubes were
transferred into liquid nitrogen (N2) after 24 hrs.
2.1.6 Recovery of Frozen Cells
Cells were removed from liquid N2 storage and thawed rapidly at 37C. 9 ml of
complete media was added to the cells in a 15 ml falcon tube and cells were pelleted at
1200 rpm for 2 mins to remove DMSO-containing media.
The cell pellet was
resuspended in 10 ml of complete media, transferred to a T25 cm2 flask and incubated at
84
the appropriate temperature in a 5% CO2 and 20% oxygen atmosphere until subconfluence was reached. Cells were then sub-cultured, as described above.
2.2 RETROVIRAL AND LENTIVIRAL INFECTIONS
2.2.1 Retroviral and Lentiviral constructs
Retroviral vector pBabepuro-wt LT cDNA and pRetroSuper were provided by O.
Gjoerup, University of Pittsburg, USA; pLPC-12SE1AORI was from S. Lowe, Cold
Spring Harbor Laboratory, USA; pBabePuro HPV16 E7 was provided by K. Munger,
Harvard Medical School, USA; pLXIPGSEp53 was provided by A. Gudkov, Roswell
Park Cancer Institute, USA; pWZLpuro-EcoR was from J. Downward, CRUK, UK;
pWZL-BlastF was from J. Morgenstern, Millenium Inc., USA; pYESir2-puro was from
Addgene and pLPCX was purchased from BD Biosciences. pLPCX-E2F-DB was
constructed by subcloning the E2F-DB gene from pCMV-DB provided by Xin Lu
(LICR, UK) into pLPCX.
The following Foxm1 constructs were obtained from Rene Medema:
pWPT-GFP: lentiviral empty vector (expressing GFP as a control for infection)
pWPT- FoxM1 wt : lentiviral human FoxM1c full-length (aa 1-763)
pWPT- FoxM1 N/KEN : lentiviral human FoxM1c N-terminal deleted (aa 210-763)
constitutively active and non-degradable (Laoukili et al., Cell Cycle 7:2720-26, 2008)
pWPT- FoxM1 R : lentiviral human FoxM1c K201, 218, 356, 460, 478, 495R)
Sumoylation-defective mutant (not published)
The FOXM1 constructs were then cloned into pLPCX by Catia Caetano.
DEPDC1, HMGB2 and MLF1-IP two splice forms Clone ID: 8860370- BC141854
(renamed 88 here) and Clone ID: 40108113- BC131556 (renamed 401 here) constructs
were purchased from geneservice and cloned into pLPCX by Parmjit Jat.
85
DBF4 (ASK) was obtained as a cDNA clone from UCL cloned into pGEM-T. It was
then subcloned into pLPCX by Parmjit Jat.
NEK2 was obtained as a cDNA clone from Andrew Fry Leicester in pGEM-3Zf(-)- 2 kb
insert and cloned into pLPCX by Parmjit Jat.
PLK4 cDNA clone was obtained from David Glover who obtained it from GeneService
and cloned into pLPCX by Parmjit Jat.
hBub1 was obtained as HA-BUB1 in pLB(N)CX blasticidin resistance vector from Ole
Gjoerup.
CDKN2C (p18) and MELK were purchased from geneservice and cloned into pLPCX
by Parmjit Jat.
LNCX-ER: RAS was kindly provided by Jesus Gil (Barradas et al, 2009).
The miR expressing vectors were purchased from Gene Service.
Lentiviral vector encoding tetracycline inducible expression of IKB-SR was provided by
P. Meier, The Breakthrough Toby Robins Breast Cancer Research Centre, UK.
Lentiviral shRNAmir silencing constructs derived from the Open Biosystems human
GIPZ lentiviral shRNAmir library, were provided by the UCL shRNA library core
facility. Lentiviral Gag/Pol expression vector p8.9 and pMDG, VSV-G viral envelope
expression vector were provided by G. Towers (UCL, UK) and D. Trono (University of
Geneva, Switzerland).
2.2.2 Viral Packaging and infections
2.2.2.1 Packaging of Retroviral Constructs
 amphotropic and  ecotropic retroviral packaging cells were plated at 1x10 6 cells/10
cm plate the day prior to transfection. Cells were transfected the following day (at
approximately 30% confluence) with 10 or 20 µg (for the micro-rna experiment) of
retroviral vector DNA and 12 µl of FuGENE 6 Transfection reagent (ROCHE),
according to manufacturer‘s instructions. 24 hrs post-transfection, media was changed
86
using 10 ml fresh media per plate. 48 hrs post-transfection, the retroviral supernatant
was harvested, filtered through a 0.45 µm filter, and either used immediately or quickly
frozen at -80ºC.
A second harvest was made when needed by re-adding 10 ml of media into the harvested
plates and harvesting the supernatant in a similar way the next day. Frozen aliquots of
retroviral supernatant were thawed rapidly at 37ºC before use.
2.2.2.2 Packaging of Lentiviral Constructs
HEK293T packaging cells were plated at 1x106 cells/10cm plate the day prior to
transfection. Cells were transfected the following day (at approximately 80%
confluence) with 1.5µg of lentiviral pGIPZ DNA vector mixed with 1µg p8.91 (gag-pol
expressor) and 1µg pMDG.2 (VSV-G expressor) and 10 µl of FuGENE 6 Transfection
reagent (ROCHE): First 200ul of medium were mixed with the fugene then after an
incubation of 5 min, the DNA mix was added and then after an incubation of 15 min, the
mix medium/fugene/DNA was added to the cells. 24 hrs post-transfection, media was
changed using 10 ml fresh media per plate. 48 hrs post-transfection, the retroviral
supernatant was harvested, filtered through a 0.45 µm filter, and either used immediately
or quickly frozen at -80ºC.
A second harvest was made when needed by re-adding 10 ml of media into the harvested
plates and harvesting the supernatant in a similar way the next day. Frozen aliquots of
retroviral supernatant were thawed rapidly at 37ºC before use.
2.2.2.3 Infection with viral supernatant and selection
Cells utilized for infection were seeded at 5x105 cells/T75 cm2 flask or 1x106 cells/T175
cm2 flasks. The following day (at approximately 30% confluence), media was aspirated,
and cells were infected with retroviral or lentiviral supernatant in the presence of 8µg/ml
polybrene. The volume of retroviral supernatant used for the infection varied for each
experiment according to the viral titre. Cells were then incubated at 34ºC for 24 hrs.
87
The following day, media was replaced with 10ml (for T-75 cm2 flasks) or 15 ml (for T175 cm2 flasks) fresh media then, 4 days post-infection, antibiotic selection was added
(2 µg/ml puromycin for retroviral infection or 6 µg/ml puromycin for lentiviral infection
or 5 µg/ml blasticidin, for the amphotropic constructs; Invitrogen), and media (including
antibiotic) was changed every 3-4 days.
2.3 DNA MANIPULATION
2.3.1 Plasmid DNA Preparation
All plasmid preparations (both small scale and large scale preparations) were carried out
using QIAGEN kits and following the manufacturer‘s instructions.
2.3.1.1 Small Scale Plasmid Preparation
Bacterial stocks were kept at -70C in LB medium containing 15% glycerol. Liquid
cultures of bacteria picked from single colonies were grown in a bacterial shaker
(vigorous shaking) overnight at 37C in 5 ml of LB medium with the appropriate
antibiotic. 1.5 ml of culture was then transferred to a 1.5 ml microfuge tube and spun at
13000 rpm for 30 sec. The cell pellet was resuspended in 250 l of solution P1 (50 mM
Tris/hydrochloric acid (HCl), pH 8.0, 10 mM EDTA and 100 mg/ml RNAse A). 250 l
of solution P2 (200 mM sodium hydroxide (NaOH) and 1% sodium dodecyl sulphate
(SDS)) was added and gently mixed by inverting the 1.5 ml microfuge tube 4-6 times.
To the same 1.5 ml microfuge tube, 350 l of solution N3 (3.0 M sodium acetate, pH
5.5) was added and immediately mixed by inverting the 1.5 ml microfuge tube 4-6
times. The mixture was then spun in a microfuge for 10 mins at 13000 rpm and the
supernatant transferred to a QIAprep column. The column was centrifuged for 30 sec at
13000 rpm then the flow-through was discarded. The column was then washed with 0.5
ml of PB buffer (QIAprep Spin Miniprep kit, QIAGEN) and then 0.75 ml of PE buffer
88
(QIAprep Spin Miniprep kit, QIAGEN). DNA was then eluted with 50 l of ddH20. All
solutions used were from the QIAfilter Plasmid Mini kit, QIAGEN.
2.3.1.2 Large Scale Plasmid Preparation
200 ml of LB medium containing the appropriate antibiotic was inoculated with an
overnight culture of bacteria and grown overnight at 37C with vigorous shaking.
Bacteria were harvested at 6100 rpm for 15 mins at 4C using an SLA 1500 rotor and
Sorvall RC5C centrifuge. The cell pellet was resuspended in 10 ml of resuspension
buffer P1 (50 mM Tris-HCl pH 8.0, 10 mM EDTA and 100 g/ml RNAse A, stored at
4C). 10 ml of lysis buffer P2 (200 mM NaOH and 1% SDS) was added and, after a 5
mins incubation step at room temperature, 10 ml of ice-cold neutralisation buffer P3
(3mM potassium acetate pH 5.5) was added and the mixture was directly applied to a
QIAfilter Cartridge. The QIAfilter Cartridge was incubated at room temperature for 10
mins before the cell lysate was filtered and directly applied to a previously equilibrated
QIAGEN-tip 500 column (equilibration buffer QBT: 750 mM NaCl, 50 mM MOPS (3 –
(N-morpholino) propanesulphonic acid) pH 7.0, 15% ethanol (v/v) and 0.15% Triton X100) and allowed to enter the resin by gravity. The column was washed twice with 30
ml of wash buffer QC (1 M NaCl, 50 mM MOPS pH 7.0 and 15% ethanol). DNA was
then eluted with 15 ml of elution buffer QF (1.25 M sodium chloride (NaCl), 50 mM
Tris-HCl pH 8.5 and 15% ethanol) and precipitated in 10.5 ml of isopropanol at room
temperature. Centrifugation was performed at 11000 rpm for 30 mins at 4C using the
SS34 rotor and Sorvall RC5C centrifuge. The DNA pellet was washed with 70%
ethanol then centrifuged again at 11000 rpm for 5 mins. The supernatant was removed
and the DNA pellet was air dried for 5 mins. DNA was resuspended in 200 l ddH2O in
a 1.5 ml microfuge tube. All solutions used were from the QIAfilter Plasmid Maxi kit,
QIAGEN.
89
2.3.2 DNA Quantification
To determine DNA concentration, the OD of the solution was measured at 260 nm
(OD260) using a Nanodrop 1000 from Thermo scientific.
DNA concentration was
calculated using the relationship: 1 OD unit at 260 nm = 50 g/ml DNA.
2.3.3 DNA-Agarose Gel Electrophoresis
DNA fragments were loaded with 1x DNA loading buffer (2.5% Ficoll, 0.04% (w/v)
bromophenol blue and 0.04% Xylene) and fractionated by electrophoresis on 1% (w/v)
agarose (Invitrogen) gels, prepared in 1x TAE (40 mM Tris-acetate and 2 mM EDTA)
with 1 g/ml ethidium bromide (BDH). Electrophoresis in 1x TAE was carried out in
electrophoresis tanks and DNA fragments were separated at a constant voltage of 100
Volts (V) for a minimum of 20 mins. Samples were loaded alongside 5 l 1kb+ DNA
ladder (Invitrogen). Ethidium bromide stained DNA fragments were then visualised on
a UVP (Dual intensity UV trans-illuminator), and an image was produced and printed
with a Sony video graphic printer.
2.3.4 DNA Sequencing
For the miniprep DNA samples, DNA sequencing was outsourced to a sequencing
service in the Institute of Neurology, Prion Unit.
2.3.5 Cloning of PCR Products
The cloning of the constructs used for complementation was the work of Dr. Louise
Mansfield or Prof. Parmjit Jat.
90
2.4 RNA MANIPULATION
2.4.1 RNA Isolation
CL3EcoR and HMF3A cultures grown in T-75 cm2 flasks were fed with fresh media the
day prior to RNA extraction and were harvested at no greater than 80% confluence on
the day of RNA extraction. Media was removed and the cells were washed twice with
1x PBS. 2.5 ml of TRIzol (Life Technologies) was then added to each T75 cm2 flask
and cells were left to lyse for 5 mins at room temperature. Cell lysates were then passed
several times through a 5 ml pipette, after which the samples were incubated for 5 mins
at room temperature. 0.2 ml of chloroform (per 1 ml of TRIzol used) was then added,
and the samples were vigorously shaken by hand for 15 sec, followed by a 5 mins
incubation step at room temperature. Samples were then centrifuged at 11000 rpm for
15 mins at 4C using a SS34 rotor and Sorvall RC5C centrifuge.
Following
centrifugation, the aqueous phase of the mixture was transferred to a fresh tube and
RNA precipitated with propan-2-ol (0.5 ml for each 1 ml of TRIzol used). Samples
were incubated at room temperature for 10 mins then centrifuged at 10000 rpm for 10
mins at 4C. Supernatant was removed and the pellet was washed once with ethanol
diluted to 75% in DEPC (diethyl pyrocarbonate) treated H 20 (0.1% DEPC dissolved in
ddH20; 1 ml of ethanol for each 1 ml of TRIzol used). The RNA pellet was briefly airdried then resuspended in 50-100 l of DEPC treated H20 and incubated at room
temperature for at least 30 mins to ensure it was completely resuspended.
2.4.2 RNA Quantification
After RNA extraction, optical density of the solution was measured at 260nm (OD260)
using a Bio-Rad spectrophotometer (Bio-Rad Smart SpecTM 3000 Spectrophotometer).
RNA concentration was calculated using the relationship:
1OD unit at 260 nm = 40 g/ml RNA
91
2.5 PROTEIN ANALYSIS
2.5.1 Preparation of Total Protein Extracts
HMF3A cultures grown in T75 cm2 flasks were fed with fresh media the day prior to
lysis and were harvested at no greater than 80% confluence on the day of lysis. For lysis,
cells were washed twice with cold 1x PBS, and 0.5 ml of 1x radioimmunoprecipitation
(RIPA) lysis buffer (150 mM NaCl, 1% Triton-X-100, 0.5% sodium deoxycholate, 0.1%
SDS and 50 mM Tris pH 8.0) was added to each T75 cm2 flask. 2 µl of Protease
Inhibitor Cocktail (2 mM 4-[2-aminoethyl] benzenesulphonyl fluoride [AEBSF], 1 mM
EDTA, 130 M Bestatin, 14 M E-64, 1 M Leupeptin and 0.3 M Aprotinin; Sigma)
was added per 1 ml of lysis buffer used. Cells were incubated on ice for 30 mins then
scraped and transferred to a 1.5 ml microfuge tube. Lysates were passed three times
through a 21-gauge needle to shear the DNA then centrifuged at 10000 rpm for 30 mins
at 4C. The supernatant from each lysis reaction was transferred to a fresh 1.5 ml
microfuge tube, aliquoted then stored at -20C.
2.5.2 Determination of Protein Concentration
Protein concentrations were determined using the Bio-Rad protein assay (Bio-Rad
Laboratories), a protein assay based on the Bradford assay [Bradford, 1976]. The dye
reagent was diluted 1:5 in PBS. A BSA standard curve was established with protein
dilutions ranging from 1-15 g/ml. 2 l of each sample were mixed with 1 ml of freshly
diluted dye and incubated at room temperature for 5 mins. OD595 was measured (BioRad Smart SpecTM 3000 Spectrophotometer) and plotted against protein concentration
of standards.
The regression coefficient was calculated and the unknown sample
concentrations determined.
92
2.5.3 Sodium-Dodecyl-Sulphate-Polyacrylamide-Gel-Electrophoresis
8 % Sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) gels
were prepared from a 30% (w/v) acrylamide stock solution (containing a ratio of 29.2
acrylamide: 0.8 N,N‘-methylenebisacrylamide; Genomic Solutions) in 375 mM TrisHCl pH 8.8 and 0.1% (w/v) SDS. Gels were polymerised by addition of ammonium
persulphate
(APS)
(0.1%
[w/v]
final;
Bio-Rad)
and
TEMED
(N,N,N‘,N‘
tetraethylenemethyldiamine, 0.0006% [w/v] final; BDH Laboratory). For 10, 12 and
15% SDS-PAGE gels, quantities of the polymerising agents were adjusted accordingly.
30 g of each cell lysate (unless otherwise stated) was heated at 90C for 5 mins with 2x
Laemmli sample buffer (8% SDS, 40% glycerol, 20% 2-mercaptoethanol, 0.008%
bromophenol blue and 0.260 mM Tris-HCl, pH 6.8) and fractionated by SDS-PAGE.
Electrophoresis was carried out at a constant voltage of 100-150 V during the day (or 40
V overnight) in running buffer (25 mM Tris, 190 mM Glycine, 0.1% [w/v] SDS).
Proteins were stacked through 2 cm of stacking gel (5% polyacrylamide, 125 mM TrisHCl pH 6.8 and 0.1% [w/v] SDS, polymerised by addition of APS and TEMED, as
before).
Proteins were fractionated alongside broad-range pre-stained SDS-PAGE
standards (Bio-Rad Laboratories).
2.5.4 Western Blotting of SDS-PAGE
Following separation via SDS-PAGE, proteins were transferred to a nitrocellulose
membrane, Hybond-c extra (Amersham Life Science) by electrophoretic transfer in a
wet tank blotting system (Bio-Rad Laboratories Trans-Blot cell). The transfer was
carried out in transfer buffer (25 mM Tris, 190 mM glycine and 20% [v/v] methanol) for
4 hrs at a constant voltage of 60 V at 4C or, alternatively, overnight at a constant
voltage of 20 V (4C).
93
The nitrocellulose membrane was then blocked by incubation in 5% (w/v) skimmed
milk powder (Marvel, Premier Brands) and 0.005% (v/v) Tween-20 (BDH Laboratory)
in 0.5x PBS (PBS/Marvel) at room temperature for 1 hr or overnight at 4C. The filter
was then incubated for 1 hr at room temperature, or overnight at 4C with the primary
antibody diluted in PBS/Marvel at the indicated dilutions (as described below). The
filter was then washed three times (15 mins each at room temperature) in 0.05% (v/v)
Tween 20 and 0.5x PBS (PBS/Tween) prior to incubation with horseradish peroxidase
(HRP)
conjugated
secondary
antibody
(Amersham
Life
Sciences
enhanced
chemiluminescence [ECLTM] western blotting analysis system) diluted 1:2000 in
PBS/Tween for 1 hr.
Following three further washes (15 mins each at room
temperature) in PBS/Tween, the filters were developed in HRP detection reagents for 90
sec, according to manufacturer‘s instructions (ECLTM, Amersham Pharmacia Biotech).
The membrane was then wrapped with Saran-wrap and exposed to an auto-radiographic
film for times varying from 10 sec to 2 hrs (Fujifilm Super RX X-ray film). Films were
developed with an AGFA X-ray film processor.
2.5.5 Antibodies Used
Anti-HPV16 E7 mouse monoclonal antibody (clone 8C9) was purchased from Zymed
Laboratories Inc, anti-cyclin D1 mouse monoclonal (clone A12), was purchased from
Santa Cruz Biotechnology; anti- Actin mouse monoclonal (clone AC-40) and anti-Tubulin mouse monoclonal antibodies (clone 2-28-33) were purchased from Sigma;
anti-p21 mouse monoclonal antibody (clone SX118) was a kind gift from X. Lu (LICR,
London); anti-E1A mouse monoclonal antibodies (clone M3 and M73) were a kind gift
from E. Harlow (Massachusetts General Hospital Cancer Center, Charlestown);
The antibodies were diluted for Western blot analysis as follows: p21CIP1/WAF1/Sdi1
(SX118) 1:500; -Actin (AC-40) 1:2000; -Tubulin 1:2000; cyclin D1 (A12) 1:1000; E7
1:100; and HRP-conjugated secondary antibodies 1:2000;
antibody M73 was used at 1:50.
94
E1A mouse monoclonal
2.6 GROWTH CURVES
2.6.1 Cell line
The cells used for this experiment are HMF3A cells. These are a conditionally
immortalized human mammary fibroblast cell line constitutively expressing U19 tsA58
LT and hTERT (O'Hare, Bond et al. 2001). These cells exhibit an immortalized
phenotype at 34ºC but undergo an irreversible growth arrest after 5 days at 38ºC. These
cells were also engineered to express the murine ecotropic receptor in order to infect
them with ecotropic viruses. Cells from a HMF3AEcoR mixed population and 6 different
clonal cell lines were used: clone #2, #3, #10, #27, #32, #33.
2.6.2 Protocol
Cells were seeded at 5000 cells per well in a 6-well plate format at day minus 1 and
grown at three different temperatures: 34ºC, 37ºC and 38ºC for 12 days. These cultures
were in duplicates for each condition (cell line and temperature) and for each time point
(126 wells in total). The numbers of cells were counted at day 0, day 5, day 7 and day 12
at these 3 growth temperatures with the cells.
2.7 IRREVERSIBILITY ASSAYS
2.7.1 Cell line
The cells used for this experiment are HMF3AEcoR cells from both the mixed population
and clonal cell line: clone #3. The passage used at the beginning of the experiment for
the mixed population is p22+11 and for the clone #3 is p22 +12 (11 and 12 passages
respectively after EcoR introduction and clone selection).
95
2.7.2 Protocol
Cells were seeded at 0.3x106 in T-75cm2 flasks at day -1 and grown at three different
temperatures: 34ºC, 37ºC and 38ºC for 7 days then at 34ºC for another 2 weeks. The
numbers of cells were counted at day 0, day 7, day 14 and day 21 at these 3 growth
temperatures with the cells being reseeded each time at 2x10 5 per T-75 cm2 flask at day
7 and 14. The reason for this reseeding is to eliminate the cooperative bias due to the cell
density. Each growth condition for each time point was represented in triplicate for
counting. Duplicate cultures were seeded at 1, 3, 5, 10, 15 and 20,000 cells per well in 6well plates and incubated at 34°C or 38°C for 7 days and then shifted back to 34°C for
14 days. Cells were stained after 3 weeks with 2% (w/v) methylene blue in 50% (v/v)
ethanol and each condition for each time point was photographed by phase-contrast
microscopy.
2.8 GROWTH COMPLEMENTATION ASSAYS
2.8.1 Cell line
The cells used for this experiment are HMF3AEcoR and CL3EcoR used at p22+11 and p22
+12 respectively.
2.8.2 Complementation experiment
The cells were seeded at 1 x 106 cells per 175 cm2 flask. The cells were then infected
with 10ml of Phoenix Ampho packaged virus supernatant for: 10µg of pLPCX, 10µg of
pLPC LT WT, 10µg of pLPC E2F-DB, 10µg of pLPC 12S E1A, 10µg of pLPC E7,
10µg of pRS Lamin A/C, 10µg of pRS p53 RNAi, 10µg of pRS p21CIP1/WAF1/Sdi1 RNAi
and 10 µg of pLXIP GSE p53 individually and 20ml of Phoenix Ampho packaged virus
supernatant for 10µg of pLPC E7 (two harvests). The cells were incubated at 34°C
96
overnight and the media was changed the next day. On day 4, puromycin selection at
2µg/ml was applied for 7 days. The culture were then washed with PBS, trypsinated,
counted and seeded at either 1K, 3K, 5K, 10K, 30K and 50K per well in 6-well plates
whenever possible in duplicate or at 0.5.105 in T-75 cm2 flasks whenever possible in
triplicate. The plates or flasks were incubated at 34°C overnight and the next day, the
media was replaced by fresh media (2ml in wells or 10 ml in T-75 cm2 flasks) before the
cells were shifted to 38°C for 3 weeks.
At week 1, 2 and 3, the cells were photographed under a microscope and at week 3, the
cells were stained with methylene blue and photographed.
2.9 SENESCENCE SPECIFIC EXPRESSION PROFILING
2.9.1 Cell line
The cells used for this experiment were CL3E coRat the passage p22+6 (6 passages after
EcoR introduction and clone selection).
2.9.2 RNA preparation
To perform the microarray procedure, total RNA was extracted from CL3 EcoR cells
incubated at 34C for 7days to prepare the reference RNA sample or at 38C for 7 days
with the various constructs described in chapter one (PLPCX, PLPC E7, PLPC E1A,
PLPC E2F-DB, PLXIP GSE p53, pRS Lamin A/C shRNA, pRS p21 shRNA and pRS
p53 shRNA) to prepare the different conditions samples to analyse. Additional total
RNA was extracted from quiescent CL3EcoR cells and HMF3S cells to prepare the
quiescence and heat shock samples. RNA was extracted from biological triplicate
cultures using Trizol (Invitrogen), frozen and sent for Analysis to the Memorial SloanKettering Cancer Center in New York.
97
2.9.3 RNA expression profiling
Expression profiling was carried out by the Memorial Sloan-Kettering Cancer Center
Core facility using Affymetrix U133 plus 2 chips. The raw expression profiling data was
then anlysed by Holger Hummerich averaged and normalised using the RMA algorithm
(Irizarry, Hobbs et al. 2003). Differentially expressed genes were identified using Linear
Models for Microarray Analysis (LIMMA). LIMMA applies a modified t-test to each
probe set employing an empirical Bayes approach for estimating sample variances. The
P-values were corrected for multiple-testing using the Benjamini-Hochberg correction
and a corrected P-value threshold of 10 -5 was used to identify significantly differentially
expressed genes.
2.10
SENESCENCE SPECIFIC MIRNA EXPRESSION PROFILING
2.10.1 Cell line
The cells used for this experiment are HMF3AEcoR mixed population cells with murine
ecotropic expression used at p22 +10.
2.10.2 Tissue culture
The cells have been cultured either at 34°C for 2 days (namely 34 samples) or at 38°C
for 7 days (namely 38 samples) DMEM supplemented with 2 mM glutamine, 100
units/ml penicillin, 100 g/ml streptomycin and only 10% v/v heat inactivated FCS or
serum-starved at 34°C for 7 days (namely 34°C quiescent samples) in DMEM
supplemented with 2 mM glutamine, 100 units/ml penicillin, 100 g/ml streptomycin
and only 0.25% v/v heat inactivated FCS.
98
For the 34°C samples and the 34°C quiescent samples the cells were seeded at 1x 10 6
cells per plate in six individual 10cm plates. For the 38°C samples the cells were seeded
at 4x 105 cells per plate in six individual 10cm plates. The cells were then grown at the
temperature, in the medium required and for the required time, as described above,
before the RNA was extracted. Each condition was grown in 6 plates to enable to pool 2
plates for 3 triplicates.
2.10.3 RNA preparation
The RNA, 9 samples in total, from the 3 condition each in triplicate was extracted, from
the 10 cm plates using miRNeasy Mini Kit from Qiagen as recommended by LC
Sciences (catalogue number 217004).
Cells from 10 cm plates were lysed directly on the plates by aspirating the cell-culture
medium, washing the cells with PBS and then adding 700μl QIAzol Lysis Reagent
(miRNeasy Mini Kit, Qiagen). The lysate was collected into a microcentrifuge tube and
vortexed to mix and ensure that no cell clumps were visible. Tubes were then frozen at 20°C until all ready to be processed together. When ready, tubes were thawed at room
temperature (15 to 25°C) for 15min and 140 μl chloroform was added to each of the
tubes containing the homogenate. The tube were shaken vigorously for 15 s and
incubated for 2-3min at room temperature. The tubes were then centrifuged for 15 min at
12,000 g at 4°C. After centrifugation, the samples separate into 3 phases: an upper,
colourless, aqueous phase containing RNA; a white interphase; and a lower, red, organic
phase. The upper aqueous phases were each transferred to new reaction tubes and 1
volume of 70% ethanol was added and mixed thoroughly by vortexing. The mixes were
transferred into an RNeasy Mini spin columns (miRNeasy Mini Kit, Qiagen) placed in a
2 ml collection tube. The tubes were centrifuged at 8000 g for 15s at room temperature
(15–25°C). The flow-throughs were then transferred into a 2 ml reaction tube and 450 μl
of 100% ethanol (0.65 volumes) were added and mixed thoroughly by vortexing. The
samples were transferred into an RNeasy MinElute spin columns (miRNeasy Mini Kit,
99
Qiagen) placed in a 2 ml collection tube. The tubes were centrifuged for 15 s at 8000 g
at room temperature (15–25°C). The flow-throughs were discarded. 500 μl of Buffer
RPE (miRNeasy Mini Kit, Qiagen) was added into the RNeasy MinElute spin columns
and the tubes were centrifuged for 15 s at 8000 g. The flow-throughs were discarded.
500 μl of 80% ethanol was added to the RNeasy MinElute spin columns and the tubes
were centrifuged for 15 s at 8000g. The flow-throughs were discarded.
RNeasy
MinElute spin columns were placed into new 2 ml collection tubes and a last spin to
rinse completely the wash buffer was performed for 5 min at 8000g. The RNeasy
MinElute spin column were placed into new 1.5 ml collection tubes and the miRNAenriched fractions were eluted by adding 14 μl RNase-free water and centrifuging for 1
min at 8000 g to elute.
2.10.4 Quality Control of RNA Samples and shipping to LC sciences
The 260 nm/230 nm ratio of each sample was analyzed by Nanodrop. The ratio should
be greater than 1.0 and the 260 nm/280 nm ratio should be above 1.8. Prior to shipping,
the total RNA was stabilised by adding 1/10th volume of 3M NaOAc, pH 5.2, then 3
volumes of 100% ethanol. The samples were then stocked at -80C until shipment.
Samples were shipped on dry ice in 1.5 ml eppendorfs.
2.10.5 Microarray analysis
2.10.5.1
Pairwise Comparisons
Microarray Analysis
for
micro-rna
Dual hybridisation is used in the LC Sciences microarray set-up so the experiment was
designed to make pairwise comparisons of the samples, whilst minimising the number of
chips required (Chapter 6, Table 6.1).
100
This service is offered by LC Sciences as ‗Total RNA to Data Service – Dual sample
option‘. In this process, the total RNA samples are enriched for miRNAs and then
labelled with Cy3 or Cy5 fluorescent dyes before hybridizing them to the same chip.
2.10.5.2
Expression analysis and normalization
This analysis is designed to identify senescence-specific miR expression by determining
which miRs are differentially expressed upon the shift from 34°C to 38°C but do not
change significantly upon quiescence or by heat shock.
The microarrays were analyzed by LC sciences: the quality of the triplicates was
checked and a normalization of the results was performed. In brief, the background is
subtracted and then signals are normalized using a LOWESS filter (Bolstad, Irizarry et
al. 2003). For the two-Cy3 and 5 dye experiments, the ratio of the two sets of detected
signals (log2 transformed, balanced) and p-value of the t-tests are calculated.
Differentially detected signals are those with p-values less than 0.01. The differential
expression between 34°C samples and 38°C samples (comparison A), and 34° and
quiescent (comparison B) was analyzed and the selection of miRNAs differential for the
comparison A but not B was achieved to create a list.
2.10.6 Individual miRNA validation in vitro
GeneService provides a micro-rna library of miRs cloned into a retroviral expression
vector and placed under control of a CMV promoter. The clones were generated in the
Netherland Cancer Institute (NKI) and made publically available by GeneService
(Voorhoeve, le Sage et al. 2006). These clones were cloned as ~500bp fragments from
several tumour cell lines. Therefore, clusters of miRNAs may be represented in an
individual miR clone (named miR-VEC).
101
2.10.6.1
Candidates miR-Vec Clones
16 MiR-Vec clones were obtained for candidate miR-18a, miR-130b, miR-372, miR373, miR-92b, miR-15a, miR16, miR-25 miR-195, mir-218, mir-20b, mir-29b, mir-186,
MiR-128, Let-7g, MiR-423-5p. In addition, hsa-let-7a1 was obtained as a negative
control. Note, hsa-miR-373 and hsa-miR-372 clones were also obtained due to the cell
transformation effects observed by Voorhoeve et al (2006).
2.10.6.2
Preparation
checking
of
clones
and
sequence
Clones were ordered from Gene Service, streaked onto an Ampicillin LB Agar plate and
a single colony was prepped up to Maxiprep level. Each maxiprep was sequenced using
T7 promoter (this sequencing was performed by MWG service as internal sequencing
was unable to read through the complex secondary structure).
2.10.6.3
HMF3A Growth Complementation Assay in
plates
20 µg of each of the 16 plasmids was packaged in 10 cm plates (in duplicate) using
Phoenix Ampho cells. Two constructs were known to be interesting candidates in cell
transformation (Voorhoeve, le Sage et al. 2006): MirVec hsa-miR-372 and MirVec hsamiR-373. A negative control, MirVec hsa-let-7a, (its expression doesn‘t vary between
34°C, 38°C or quiescence) and a positive control, pRS p21F, were also packaged for the
same experiment.
HMF3A cells (at passage p26) were infected with 50 ml of each amphotropic retrovirus
(5 harvests) then selected with 5 µg/ml blasticidin for 15 days or puromycin at 2 µg/ml
for 4 days (for the p21CIP1/WAF1/Sdi1 RNAi construct). Cells were reseeded in a 6-wellplate format with 1, 3, 5, 10, 30 and 50K cells (in duplicate, when possible) or at 0.5x105
or 1x105 or 1.2x105 (in triplicate when possible), then shifted to 38°C for 3 weeks.
102
The cultures of each condition were stained at the end of the 3 weeks with methylene
blue dye (2% (w/v) methylene blue in 50% ethanol and ddH2O) and photographed to be
analysed.
2.11
SHRNA SCREENING
2.11.1 Cell line
The cells used for this experiment were CL3E coR at the passage p22+6 (6 passages after
EcoR introduction and clone selection).
2.11.2 RNAi library
The RNAi library consists of 100 tubes of plasmids pools each containing between 150
to 200 different shRNA plasmids. Each gene is represented by 1 to 3 shRNA plasmids
and each plasmid is complementary to a different region of the target gene. Multiple
shRNA plasmids per gene are used in order to increase the likelihood of achieving
maximum efficiency of gene knockdown. The library represents 20,000 constructs to
test or 10,000 genes targeted.
2.11.3 Virus packaging
Cells were seeded at 1 x 106 in 10cm plates (day 0) and transfected the next day (day 1).
For the transfection, 12µl Fugene transfection reagent was added into 100µl media and
the mix was incubated at room temperature for 5 minutes. Then, 10µg of plasmid DNA
pool was added to the mix, and mixed gently by tapping. The mix was incubated at room
temperature for 15-30 minutes. Prior to addition to the cells, the mix was pipetted up and
103
down to mix. It was then added dropwise over the cells. The media was changed on day
2 (12ml) and the supernatant was harvested the next day (day 3). The supernatant was
then filtered through a 0.45µm membrane to remove any cells and aliquoted into 2x 1ml
for virus titration and 10ml for the screen. Virus supernatants were frozen at -70C or
used immediately. 12ml of media was added to the cells for a second harvest the next
day (day 4), as described above.
2.11.4 Sensitivity of the model
A mixture was created by mixing a quantity of positive pRS p21F RNAi constructs at
1/200 with negative pRS Lamin A/C constructs. This spiked mix was packaged in
Phoenix Eco cells and used to infect CL3EcoR at 0,5x106 in a T-75 cm2 flask (day 0).
Along with it, a positive control, P21F RNAi and a negative, Lamin A/C constructs were
each packaged and used to infect a flask of cells. The media was changed the next day
and puromycin selection at 2µg/ml was added on day 4. At day 8, the cells were
trypsinated and reseeded at 8,5x104 per 15cm plate or 0,5x104 per well in 6-wells plates.
The next day (day 9) the media was changed and cells were shifted to 38°C for 3 weeks.
At that point, the cells were stained using methylene blue dye.
2.11.5 Titration of Phoenix Eco viral Supernatants
Cells were seeded at 6x104 cells per well in 6-well plates (day 0) and infected the next
day (day 1) with different volumes (from 0.5 ml to 1x10 -4 ml of each virus pool in
presence of 8μg/ml polybrene. The media was changed on day 2 and puromycin
selection at 2 µg/ml was added on day 4. After 2 weeks, puromycin selection at 34°C,
the cells were stained with methylene blue and the number of colonies counted. The
volume required to obtain approximately 10,000 infectious events was determined.
104
Unfortunately, because the amount of DNA available to us was limited, for the pools
with a low titer, the volume of viral supernatant used was set at 10 ml (maximum
amount harvested).
2.11.6 Experiment planning for each plasmids pools
Cells were seeded at 0.5x106 per T-75 cm2 flask on day 0 and infected on day 1 with the
determined volume of virus supernatant in presence of 8μg/ml polybrene.
Accompanying every experiment, a positive control, P21F RNAi construct virus
supernatant, and a negative control, Lamin A/C construct virus supernatant, were each
used to infect a flask of cells. The media was changed on day 2 and puromycin selection
at 2µg/ml was added on day 4. At day 8, the cells were trypsinated and reseeded at
5.3x104 per T-75cm2 flasks or 1.8x104 per T-25cm2 flask. The next day the media was
changed and cells were shifted to 38°C.
2.11.7 Confidence intervals
Using the formula: ln (1-.95) / ln (1-1/(Library Size)), recommended for genetic
screens by Nolan labs (see http://www.stanford.edu/group/nolan/screens/screens.html),
it is possible to calculate the number of essays needed depending on the size of the
library and the confidence interval wanted (Chapter 4, Table 4.1).
If the interval of confidence chosen was 99%, the number of essays would have to be
superior or equal to 43,254 for my library of 9393 genes. In the shRNA screen process,
the number of cells reseeded after puromycin selection was 5.3x10 4 which is superior to
43,254 so the confidence in the results are superior or equal to 99%.
105
2.11.8 Genomic DNA extraction
After 3 weeks at 38°C, the cells were trypsinised and reseeded in the same T-75cm2
flask and grown to confluency. When confluency was reached, the cells were passaged
once more. When cell numbers were sufficient (90% confluency), the genomic DNA
was extracted from a near confluent T75 cm2 culture using the QIAamp DNA Mini kit
(Qiagen). Media was removed from the culture and the cell monolayer was washed
once with 1x trypsin-EDTA (0.25% (v/v) trypsin and 0.03% (w/v) EDTA). The
monolayer was then detached, using 1x trypsin-EDTA (0.5 ml/T25 cm2 flask) for 5 mins
at 34C), and the trypsin-EDTA was then inactivated by adding 1 ml of complete media.
Cells were transferred to a 1.5 ml microfuge tube and centrifuged for 5 mins at 3000
rpm. The supernatant was discarded and the cell pellet was resuspended in 1x PBS
(Phosphate Buffered Saline (without CaCl2 or MgCl2)) to a final volume of 200 l,
before 20 l of QIAGEN Protease (QIAamp DNA Mini Kit, Qiagen) was added. 200 l
Buffer AL (QIAamp DNA Mini Kit, Qiagen) was then added and mixed by pulsevortexing for 15 sec, followed at incubation at 56C for 10 mins. The 1.5 ml microfuge
tube was centrifuged briefly then 200 l ethanol (96-100%) was added, followed by
pulse-vortexing for 15 sec. The mixture was then applied to a QIAamp Spin Column
(QIAamp DNA Mini Kit, Qiagen) and centrifuged at 8000 rpm for 1 min. The QIAamp
Spin Column was placed in a clean 2 ml collection tube, washed with 500 l Buffer
AW1 (QIAamp DNA Mini Kit, Qiagen) and then centrifuged at 8000 rpm for 1 min.
The QIAamp Spin Column was placed in a clean 2 ml collection tube then washed with
500 l Buffer AW2 (QIAamp DNA Mini Kit, Qiagen) and centrifuged at 1300 rpm for 3
mins. The QIAamp Spin Column was then placed in a clean 1.5 ml microfuge tube and
200 l Buffer AE (QIAamp DNA Mini Kit, Qiagen) was added. Following incubation
at room temperature for 1 min, the 1.5 ml microfuge tube was centrifuged at 8000 rpm
for 1 min. The DNA concentration was determined on Nanodrop using 1.5 µl. The DNA
was then used for TOPO cloning before sequencing.
106
2.11.9 TOPO cloning and sequencing
200 ng genomic DNA was used in a 50 l PCR reaction that contained 1 l of 6.6 M
primers pSM2 longF and 6.6 M pSM2 longR, 1 l KOD Hot Start DNA Polymerase
(Novagen), 3 l 25 mM MgSO4 (Novagen) and 5 l 2 mM dNTPs (Novagen). An
initial denaturation step at 95C for 2 mins was performed before PCR amplification.
PCR amplification parameters were denaturation at 95C for 20 sec; annealing at 60C
for 10 sec; extension at 70C for 30 sec, and a final extension of 10 mins at 70C after
the last cycle. 40 cycles were used in total. The DNA was generally not visible after
this first round of PCR so a second round of amplification using a set of nested primers,
namely pSM2 F and pSM2 R, that were internal to the first set of primers was used to
amplify the inserts in the same condition than previously (5ul of PCR product used)
before TOPO-cloning. This time analysis by electrophoresis of 5μl of PCR product
revealed a product of 424 bp in all samples that corresponded to the expected insert
sequence, but not in a negative control sample where water was substituted for template
DNA.
A PCR reaction containing 100 ng pSM2 scrambled was used as the positive control for
PCR amplification, and a PCR reaction that contained no template was used as the
negative control to check for contamination of the reaction mixture. 15ul of the
amplified product were then resolved alongside 5 l 1kb+ DNA ladder (Invitrogen) on a
3% agarose gel to check for the generation of 438 bp PCR products that could be
visualised on a UVP.
4 l of each PCR product was directly cloned into pCR2.1-TOPO vector (Invitrogen)
using the TOPO TA Cloning Kit (Invitrogen), as described above. 4 l of the cloning
reaction was transformed onto LB-agar plates containing 50 g/ml final concentration
ampicillin and 80 l of 20 mg/ml X-gal and incubated at 37C overnight. Blue/white
selection was used to identify positive clones that were picked and prepped using the
107
QIAprep Spin Miniprep kit (QIAGEN), as described above. Positive clones were
sequenced using M13R primer.
2.12
PRIMARY CELLS IMMORTALIZATION
First, viral supernatant was packaged using Phoenix Ampho cells for each of the
following constructs: MiR-25, Mir-372, MiR-218, Mir-193b, MiR-423-5p, Mir-195, WT
LT, p53 shRNA, p21 shRNA. 20 µg of each of the miRs constructs were packaged for
10µg of the others constructs. The construct ER-RAS was also packaged with 10µg
DNA in Phoenix Ampho cells. hTERT viral supernatant was prepared from TEFLYA
TERT producer cells (O‘Hare et al, 2001).
The BJ cells were infected with 10 ml viral supernatant for hTERT and 40 ml of miR
viral supernatant. The cells were then selected with hygromycin at 50 µg/ml (for hTERT
alone) for at least 10 days and then with blastocidin at 2.5 µg/ml (for mir cultures) for at
least 8 days, for WT LT, p53 shRNA and p21 shRNA. The selection was done with
puromycin at 1µg/ml for at least 6 days. Control cells submitted to puromycin,
blastocidin or hygromycin died in respectively 4, 7 and 9 days.
Upon completion of selection, all cultures were infected with 10 ml of ER-RAS and
selected with geneticin (G418) at 0.75 mg per ml for 10 days. Control cells submitted to
G418 died in 9 days.
To ensure that ER-RAS was not activated, cells immediately after infection, were
transferred into phenol-red free medium supplemented with charcoal-stripped FBS,
because a lipophilic impurity contained in the phenol red has been described as a weak
estrogen agonist (Berthois et al, 1986).
108
3
CELL MODEL AND GROWTH COMPLEMENTATION
3.1 CREATION OF CLONAL CELL LINES DERIVED FROM HMF3A
CELLS AND GROWTH COMPLEMENTATION ASSAYS
3.1.1 Objectives
My objectives were to specify the senescence model and test the sensitivity of the
complementation assay in the mixed population of HMF3AEcoR cells. It was also to
create a refined clonal model, CL3 EcoR, to assess the sensitivity of this model, its
consistency compared to the mixed population and to confirm whether complementation
of the growth arrest with abrogation of either the p53 or the pRb pathway was able to
bypass cellular senescence in the mixed population cells. Another main objective was to
design protocols and optimise conditions to develop a more standardised assay that
results in a minimal background. In addition, investigation of a new expression
construct, namely E2F-DB, as an alternative method to abrogate the pRb pathway was
tried.
3.1.2 Refinement of the HMF3A cells by clonal selection
The nature of the complementation experiment requires a reproducible cell response in
order to be able to compare complementation with different constructs. For that purpose,
the mixed HMF3A cells were infected with a retrovirus transducing the murine
ecotropic receptor and then, after antibiotic selection, single cell clone colonies were
picked and a total of 35 different clones were selected. After growing the 35 clones, 6
fast growing clones were selected for further evaluation. The evaluation was performed
by comparing the growth rates and the irreversibility of all the clones to the mixed
population. As cellular senescence is an irreversible growth arrest, it was important for
the new clonal model to reflect this property. In addition, similar growth
109
complementation assays were performed with the clones and with the most sensitive
constructs to assess the accuracy of the model.
3.1.2.1 Growth curves
In order to choose a clone for the experiments, the growth rate of the mixed population
was compared to those of 6 different clonal cell lines of the HMF3AEcoR cells: clone
#2, #3, #10, #27, #32, #33 were chosen from among approximately 35 clones as they
were the fastest growing.
Cells were each grown at three different temperatures: 34ºC, 37ºC and 38ºC for 12 days.
The numbers of cells were counted at day 0, day 5, day 7 and day 12 in triplicate. Each
culture was seeded at day -1 at a similar density of 5000 cells per well. The cultures at
34 ºC do not have a day 12 value as the cultures were overgrown by that point and were
starting to detach. The cell numbers were very similar between the cells growing at 37°C
and at 38°C (Figure 3.1). On this Figure, the two temperatures 37°C and 38°C do not
seem to have different effects on the cell growth rate. The cell numbers are relatively
similar between clones and mixed population. However, when comparing the numbers
on a same scale, it appears that there are three categories: slow growth clones (clones 2,
10 and 33) with numbers hardly reaching 20,000 cells after 7 days; medium pace growth
clones (clones 27 and 32 and the mixed population) with numbers around 30,000 cells
and one quick growth clone: clone 3. Clone #3 seems to be the clonal line with the best
growth rate at 34ºC reaching numbers of 45,000 cells at day 7. It is also possible to note
that all the clonal cells lines undergo growth arrest at both 37ºC and 38ºC from day 5
(Figure 3.1).
110
Figure 3.1: Clonal cell lines growth rates
The growth rate of the mixed population was compared to those of 6 different clonal cell lines of the
HMF3AEcoR cells: clone #2, #3, #10, #27, #32 and #33. Cells were each grown at three different
temperatures: 34ºC, 37ºC and 38ºC for 12 days. The numbers of cells were counted at day 0, day5, day 7
and day 12 in triplicate. Each culture was seeded at day -1 at a similar density of 5000 cells per well.
111
3.1.2.2 Complementation with E7 and E1A
Dr. Louise Mansfield had shown that senescence growth arrest in the HMF3A cells can
be abrogated by ectopic expression of a certain number of genes and Ad E1A and HPV
E7 represented the least efficient. I wanted to ensure that the clonal cell line could also
be complemented by HPV E7 ectopic expression. This test would be a good evaluation
of the clonal representativity and sensitivity.
It was important to make sure that both the mixed population and the clonal cell lines in
parallel experiment could be rescued with these constructs. The first step was to infect
the cells with ecotropic retroviral supernatant of pLPCX alone and pLPC-Ad E1A,
pLPC-HPV E7, pRS p21 shRNA and p53GSE. The cells were reseeded after antibiotic
selection at 1000, 5000, 10,000, 15,000 and 20,000 cells per well in 6-well-plates and
grown at 38°C for 3 weeks before being stained (Figure 3.2). Only the three higher
densities results are shown here as they are more significant numbers to compare and as
in the lower ones almost no rescuants were observed. For these reasons, the experiment
was repeated with slighty higher densities of 15,000, 20,000, and 30,000 cells per well
(Figure 3.3).
The results show almost no background for the mixed cell culture nor for the clonal cell
lines in both experiments (Figure 3.2A and 3.3A). The results in the mixed population
cells with the E7 constructs show rescue at the higher densities of 20,000 or 30,000 cells
per well (Figure 3.2D and 3.3E) although at a very low level compared to the positive
control p21 shRNA or p53GSE (Figure 3.2C and 3.3B and C). The results are generally
better in flasks probably because there could be more stress on cells plated in 6-well
plates. The only 2 clonal lines to reproduce the mixed population results were clone #3
and clone #10, although, with clone #3, the cells rescue at a much higher level as it is
possible to see stained colonies at 15,000 cells per well (Figure 3.2D and 3.3E).
112
Figure 3.2: Growth complementation assay in the clonal cell lines
The mixed population and the clonal cell lines #2, #3, #10, #27, #32 and #33 were tested for growth
complementation in parallel experiment. The cells were infected with ecotropic retroviral supernatant of
pLPCX alone and pLPC-Ad E1A, pLPC-HPV E7, pRS p21 shRNA and p53GSE. The cells were reseeded
after antibiotic selection at 1000, 5000, 10,000, 15,000 and 20,000 cells per well in 6-well-plates and
grown at 38°C for 3 weeks before being stained
113
Figure 3.3: Growth complementation assay in the clonal cell lines repeat
The mixed population and the clonal cell lines #2, #3, #10, #27, #32 and #33 were tested for growth
complementation in parallel experiment. The cells were infected with ecotropic retroviral supernatant of
pLPCX alone and pLPC-Ad E1A, pLPC-HPV E7, pRS p21 shRNA and p53GSE. The cells were reseeded
after antibiotic selection at 15000, 20000, and 30,000 cells per well in 6-well-plates and grown at 38°C for
3 weeks before being stained
114
The results with E1A show very similar results: the mixed population displays rescued
colonies but very few and only at the higher reseeded density of 20,000 cells per well
(Figure 3.2B and 3.3D), while the only clonal cell line to rescue significantly was clone
#3 and at the densities of 15,000, 20,000 and 30,000 cells per well (Figure 3.2D and
3.3E).
The decision was made to test and validate clone #3 (CL3EcoR) as the model for this
conditional senescence system. Since CL3EcoR grew the best, growth arrested at 37 and
38°C and also rescued with both HPV E7 and Ad E1A, I chose to take it forward for
testing.
3.1.2.3 Irreversibility
Since cellular senescence is an irreversible growth arrest, it was very important to show
that CL3EcoR and the mixed population were undergoing an irreversible growth arrest
and that the CL3EcoR cells were representative of the HMF3A mixed population.
At the same time testing the irreversibility would permit the choice of temperature
conditions that would eliminate reversible growth arrest.
 Microscopical observation
Senescent cells display an increase in cell size and a flattened phenotype. To confirm if
the HMF3AEcoR mixed population and CL3EcoR displayed these morphologic
characteristics, senescence was triggered in the conditional model and the cells were
observed microscopically. The conditionality of the mixed HMF3A EcoR cells was tested
by growing the cells at 34°C, 37°C or 38°C for 7 days and then back at 34°C for another
14 days after which the cells were observed and photographed under phase contrast
microscope (Figure 3.4). The cells were also reseeded at an identical density after each
photograph at day 7 and 14 to eliminate the density bias. I observed that a high density
helped the cell growth while a low density slowed down the growth.
115
Day 0
Day 7
Day 14
Day 21
A
B
C
D
E
F
G
H
I
J
K
L
34°C
37°C
38°C
HMF3AEcoR
34.0 C
M
38.0 C
N
CL3EcoR
Figure 3.4: Irreversibility HMF3AEcoR (A-L) and CL3EcoR (M-N) cells: Photos
Irreversibility was tested in 6-well plates by incubating cells at 34°C or 38°C for 7 days and then shifting
them back to 34°C for 14 days before photographing the cells under phase contrast microscope.
116
This density effect has also been previously described in the literature (Piedimonte,
Borghetti et al. 1982). Cells all clearly display a flattened phenotype after 7 days at 37°C
or 38°C (Figure 3.4, F and J). After shifting the cells back to 34°C, however, although
the growth does not restart, the cells all display again a much smaller size and lose their
flattened senescent phenotype (Figure 3.4, G-H and K-L). It is also possible to note that
the density is much greater in the cells grown at 34°C all along (Figure 3.4, A-D)
indicating a growth arrest with minimal background at the higher temperatures (Figure
3.4, E-L). The growth rate at 34°C seemed consistent all the way through. The cells at
day 21 when grown at 34°C appeared slightly less dense on the photograph (Figure 3.4
D); however, the culture had similar cell numbers than the other days and the visual
difference was only due to which area was photographed on the flask. There was no
visible differences between the cells grown at 37°C or at 38°C for 7 days (Figure 3.4, F
and J), however, after shifting them back to 34°C, it appeared that the cells grown at
37°C started growing again (Figure 3.4, G and H) when compared to the ones grown at
38°C (Figure 3.4, K and L). These results suggested that 37°C was not stringent enough
to trigger an irreversible arrest and for that reason, I chose to perform the
complementation assays at 38°C.
In addition, the growth arrest of Clone #3 (CL3 EcoR) was tested by growing the cells for
7 days at either 34°C or 38°C. Microscopical observation showed healthy growing cells
at 34°C and arrested cells with a similar flattened phenotype to the one of the mixed
population at 38°C (Figure 3.4, M-N).
 Do Cell numbers confirm microscopical observation?
The mixed population of HMF3AEcoR cells were each plated at 0.3x106 per T-75 cm2
flask (day 0) and grown at 34°C or 38°C for 7 days (day 7) and then at 34°C for another
14 days (day 14 and 21). The numbers of cells were determined at day 0, day 7, day 14
and day 21 at these 3 growth temperatures with the cells being reseeded at 2x10 5 per T75 cm2 flask at day 7 and 14. Reseeding eliminates the potential bias due to different
cell densities in cultures. Each growth condition for each time point was represented in
117
triplicate. These cells numbers were represented as the accumulated relative growth at
each time point (Figure 3.5A). The cells at 34°C for 3 weeks showed a growth rate just
above 20 divisions per week. The cells grown at 38°C on the other hand seemed to have
their growth arrested. Even when the cells were shifted back to 34°C for 2 weeks, the
growth remains very poor as their relative growth is only around 5 divisions per week.
This residual growth when the cells were shifted back to 34°C could be due to a low
number of reversions, even though very little background is observed when the cells are
maintained at 38°C. Together, these data suggest that the cells undergo irreversible
growth arrest. Before concluding that this growth arrest was a senescence growth arrest,
more tests were performed.
 Cell staining
To confirm the microscopic results, HMF3AEcoR and CL3EcoR cells were plated at
different densities in 6-well plates and grown at 34°C or 38°C for 7 days and then at
34°C for another 14 days. Cells were then stained with methylene blue and the plates
scanned (Figure 3.5, B and C). The methylene blue dye used here stains healthy growing
cells as dark blue. The flattened cells are stained in a much lighter shade. Cells from
both HMF3AEcoR and CL3EcoR cultures show a greater intensity of staining for the 34°C
samples than the 38°C ones, indicating no growing healthy cells at 38°C. This suggested
the HMF3AEcoR and CL3EcoR cells undergo growth arrest at 38°C that is essentially
irreversible. It is interesting to note that at the density of 10,000 cells or lower, there is
very little background suggesting an appropriate density to use for the complementation
assay.
118
Figure 3.5: Irreversibility HMF3AEcoR (A and B) and CL3EcoR (C) cells: Growth assay and
staining
Irreversibility was tested in 6-well plates by incubating cells at 34°C or 38°C for 7 days and then shifting
them back to 34°C for 14 days before staining (B and C).
It was also determined by counting
HMF3AEcoR cells numbers achieved after culturing cells at 34°C or 38°C for 7 days and then at 34°C
for another 14 days (A).
119
 SA-β-Gal activity
Senescence represents an arrested state in which the cells remain viable, but are not
stimulated to divide by serum or passage in culture. Senescent cells display increase of
cell size which has been confirmed by microscopy but also display activation of
senescence-associated expression of β-galactosidase (SA-β-Gal) activity and altered
patterns of gene expression. The SA-β-Gal activity is detectable by a histochemical
reaction only in senescent cells and is not found in pre-senescent, quiescent or immortal
cells. In order to assess whether the CL3EcoR cells were undergoing senescence upon
growth arrest, cells were plated in 6 well plates at 5000 cells per well and grown at 34 or
38°C for 7 days. The cells were then stained for SA-β-galactosidase activity (Figure 3.6,
C and D). The same experiment was performed in parallel with HMF3S cells. The
HMF3S cells were derived from the same batch of primary human breast fibroblasts by
immortalisation with hTERT and wild type SV40 U19 LT antigen and thus did not
undergo growth arrest upon shift of temperature (Figure 3.6, A and B).
The results showed no blue colouration and healthy growing phenotype of the HMF3S
cells at 34°C and 38°C (Figure 3.6, A and B). There is no blue colouration either of the
CL3EcoR cells at 34°C (Figure 3.6C). At 38°C, the CL3EcoR cells show an intense blue
colouration situated in the cytoplasm of the cells (Figure 3.6D). The cells were flattened
and displayed a senescent phenotype. Thus, CL3EcoR cells undergo irreversible growth
arrest at 38°C. This growth arrest at 38°C was stringent, essentially irreversible and
turned on SA-β-Gal activity, a marker of senescence.
Complementation assays by abrogation of the pRb and the p53 pathways were
performed in the CL3EcoR cells and the mixed population.
120
Figure 3.6: Induction of SA-β-Gal
CL3EcoR and HMF3S cells were incubated at either 34°C or 38°C for 7 days and were stained for SA-βGal activity. HMF3S cells which do not undergo growth arrest at 38°C were analysed as a temperature
control.
121
3.1.3 Reconstitution of WT LT antigen in HMF3AEcoR and CL3EcoR cells
Reconstitution of WT LT activity in HMF3A cells by infection with amphotropic
viruses had been shown by Dr. Louise Mansfield to be suficient to abrogate growth
arrest. This experiment was carried out in the new conditions with ecotropic virus in the
CL3EcoR and the mixed population HMF3AEcoR Cells.
10 g of pLPC WT LT and empty pLPCX VECTOR were packaged, using the 
ecotropic cells. 10ml of the retroviral supernatant was used to infect CL3 EcoR and
HMF3AEcoR cultures seeded at 5x105 cells in duplicate T75 cm2 flasks in the presences
of 8 g/ml polybrene. Following incubation at 34C for 4 days, 2 g/ml puromycin was
added to the culture medium and, after completion of 4 days of drug treatment, no viable
cells remained in a non-infected culture, whereas multiple puromycin-resistant clones
were observed in all infected cultures.
Selection was removed and the cells were
reseeded in 6 well-plates at 1000, 3000, 5000, 10,000, 30,000 and 50,000 cells per well
before being shifted to 38C for a further 14 days before fixing and staining with 2%
(w/v) methylene blue (Figure 3.7 and 3.8).
The staining result showed no or very little background with the pLPCX empty vector
(Figure 3.7A and 3.8A) which meant that the cells were unable to overcome growth
arrest on their own but showed growth with the WT LT vector (Figure 3.7B and 3.8B) at
all densities confirming that reconstitution of WT LT was sufficient to overcome the
growth arrest in the CL3EcoR and HMF3AEcoR cells in a similar manner.
Western blot analysis of WT LT expressing mixed population HMF3A EcoR cells by
western blot with p21CIP1/WAF1/Sdi1 antibody showed a decrease in the expression of
p21CIP1/WAF1/Sdi1 (Figure 3.9B). Indeed, LT is known to bind and inactivate several
proteins including pRb and p53 and p53 is directly upstream of p21CIP1/WAF1/Sdi1.
Therefore, the inactivation of p53 by LT should trigger a decrease in the p21CIP1/WAF1/Sdi1
proteins levels (Figure 3.9A).
122
Figure 3.7: Complementation HMF3AEcoR by ectopic expression and RNAi silencing
Cells stably transduced with the retroviruses corresponding to pRS Lamin A/C (control gene), pRS p53
shRNA, pRS p21F shRNA, HPV-E7, E2F-DB, 12S E1A, PLPCX, WT LT and p53 GSE were seeded in 6well plates at 1000, 3000, 5000, 10000, 30000 and 50000 and incubated at 38°C for 21 days before
staining. Constructs able to overcome growth arrest yielded dark blue colonies of densely growing cells.
123
Figure 3.8: Complementation CL3EcoR by ectopic expression and RNAi silencing
Cells stably transduced with the retroviruses corresponding to pRS Lamin A/C (control gene), pRS p53
shRNA, pRS p21F shRNA, HPV-E7, E2F-DB, 12S E1A, PLPCX, WT LT and p53 GSE were seeded in 6well plates at 1000, 3000, 5000, 10000, 30000 and 50000 and incubated at 38°C for 21 days before
staining. Constructs able to overcome growth arrest yielded dark blue colonies of densely growing cells.
124
Figure 3.9: Expression of LT in the HMF3AEcoR cells
At 38°C, the thermolabile LT is inactivated, p53 gets activated leading to the increase of p21CIP1/WAF1/Sdi1
expression directly downstream (A). If LT expression is reintroduced, p53 is inactivated and
p21CIP1/WAF1/Sdi1 expression levels stay low (A) The expression of p21CIP1/WAF1/Sdi1 protein was analyzed by
Western blot in cells transfected with either PLPCX or WT LT at 34 and 38°C (B).
125
3.1.4 Abrogation of the p53 pathway in HMF3AEcoR and CL3EcoR cells
Louise Mansfield had also shown that abrogation of the p53 pathway in the mixed
population of HMF3A cells by ecotropic retroviral delivery was sufficient to bypass the
conditional growth defect.
This experiment was repeated in the new conditions with the CL3 EcoR and HMF3AEcoR
cells to confirm a similar response. To inactivate the p53 pathway, three different
reagents were used, one by ectopic expression and two by shRNA silencing; namely
pLXIP GSE p53, pRS p21 shRNA and pRS p53 shRNA.
GSE p53 is a dominant-negative peptide of p53 that was originally identified in a GSE
screen. It corresponds to a region in the oligomerisation domain of p53 (amino acids
273-368 in rat) (Ossovskaya, Mazo et al. 1996) and functions as a dominant-negative
peptide of p53 by promoting the accumulation of endogenous p53 protein into a
functionally inactive form. However, the high level of sequence conservation exhibited
in the oligomerisation domain between p53 and the p53 family members p63 and p73
(Levrero, De Laurenzi et al. 2000), suggests that GSE p53 probably interacts with all
three members of the p53 family. Therefore, RNAi represented a second, more specific
method to abrogate p53 activity.
A p53 shRNA construct, pRetroSuper-p53, had previously been shown to efficiently
knockdown p53 in HDFs (Berns, Hijmans et al. 2004). Therefore, the same shRNA
construct was reconstructed by Dr.Louise Mansfield by cloning the p53 RNAi target
sequence into the pRetroSuper retroviral expression vector.
p21CIP1/WAF1/Sdi1, directly downstream of p53, has been shown to be up-regulated upon
replicative senescence in a number of cell types (Schwarze, Shi et al. 2001; Wagner,
Hampel et al. 2001; Tang, Gordon et al. 2002; Hardy, Mansfield et al. 2005).
Furthermore, over-expression of p21CIP1/WAF1/Sdi1 in HDFs was shown to induce
126
premature senescence (McConnell, Starborg et al. 1998) and SAHF (senescence
associated heterochromatin foci) (Chan, Narita et al. 2005), whereas knockdown of
p21CIP1/WAF1/Sdi1 by shRNA was sufficient to bypass the conditional growth arrest in an
analogous conditionally immortalised HDF system, namely BJ-TERT-tsLT cells. This
indicated that p21CIP1/WAF1/Sdi1 should be functionally analysed in the HMF3A system. A
number of p21CIP1/WAF1/Sdi1 shRNA constructs were designed by Dr. Louise Mansfield
using the criteria outlined by Reynolds and colleagues (Reynolds, Leake et al. 2004) to
find one that worked best in the HMF3A cells to silence p21CIP1/WAF1/Sdi1 ;pRetroSuperp21F.
10 g of pRetroSuper-p53, pRetroSuper-p21, pRetroSuper Lamin A/C control vector,
pLPC-GSEp53 and empty pLPCX vector each were packaged, using the  ecotropic
cells. The complementation assay was performed as described previously for WT LT
and the cells were stained with 2% (w/v) methylene blue (Figure 3.7 and 3.8).
Whereas no growing colonies could be observed in the control-infected cultures pLPCX
and pRS Lamin A/C shRNA (Figure 3.7 and 3.8, A and G), multiple colonies could be
observed with a blue colouration in the p53 shRNA, p21 shRNA and GSE p53-infected
cultures incubated at 38C for 21 days, incubated under the same conditions (Figure 3.7
and 3.8, C, H, I) for both HMF3AEcoR and CL3EcoR cells. This indicated that downregulation of p53 or p21CIP1/WAF1/Sdi1 was sufficient to complement the growth of these
cells under non-permissive conditions.
It is interesting to note that the efficiency of p53GSE to abrogate growth arrest was
superior to p53 shRNA for both HMF3AEcoR and CL3EcoR cells which could be due to
the contributing effects of the potential inactivation of p63 and p73. Additionally, the
p21 shRNA also abrogated growth arrest more efficiently than p53GSE and p53 shRNA
in both HMF3AEcoR and CL3EcoR cells. This could be explained by either a better
knockdown of p21CIP1/WAF1/Sdi1 by the shRNA than the p53. Similar complementation
results were obtained in a duplicate experiment.
127
Protein lysates of each condition were analysed by western blot with a p21CIP1/WAF1/Sdi1
antibody to assess whether the expression of p53 GSE, p53 shRNA and p21 shRNA
would affect the proteins levels of p21CIP1/WAF1/Sdi1.
p53 is situated directly upstream of p21CIP1/WAF1/Sdi1 and its inactivation by either p53
GSE (Figure 3.10A) or p53 shRNA (Figure 3.10C) brings p21CIP1/WAF1/Sdi1 proteins
levels down. Similarly, p21 shRNA also brings the p21CIP1/WAF1/Sdi1 proteins levels down
(Figure 3.11A).
The results show endogenous expression of p21CIP1/WAF1/Sdi1 protein in the cells with the
empty vector pLPCX or the control shRNA Lamin A/C and, as expected, an increase in
p21CIP1/WAF1/Sdi1 protein levels at 38C compared to 34C (Figure 3.10B and D, and
Figure 3.11B lane 1 and 2). When the cells express either p53 GSE or p53 RNAi, or the
p21 RNAi, the p21CIP1/WAF1/Sdi1 protein levels are considerably reduced at both 34 and
38C (Figure 3.10B and D, and Figure 3.11B, lane 3 and 4).
3.1.5 Abrogation of the pRb pathway in HMF3AEcoR and CL3EcoR cells
Similarly to p53, the pRb gene and genes that operate in the pRb pathway are frequently
inactivated in most types of human cancer, either by direct mutation of pRb itself, or by
mutation of an upstream regulator (Sherr 1996; Sellers and Kaelin 1997; Nevins 2001;
Hahn and Weinberg 2002; Ortega, Malumbres et al. 2002). However, targeting pRb for
inactivation is complicated by the problem of functional redundancy resulting from both
the multiplicity of both Rb family members and potential pRb-binding partners.
The viral oncoproteins adenovirus type 5 E1A and HPV type 16 E7 represent two
reagents commonly used to inactivate this pathway and Dr. Louise Mansfield showed
that abrogation of the pRb pathway with either E1A or E7 expression could bypass the
senescence growth arrest.
128
Figure 3.10: Expression of p53 in the HMF3AEcoR cells
At 38°C, the thermolabile LT is inactivated, leading to an increase of p53 and therefore the increase of
p21CIP1/WAF1/Sdi1 expression directly downstream (A-C). If p53 is repressed (A) or silenced (C),
p21CIP1/WAF1/Sdi1 expression levels stay low (B-D) The expression of p21CIP1/WAF1/Sdi1 protein was analyzed
by Western blot in cells transfected with either PLPCX, p53 shRNA or p53 GSE at 34 and 38°C (B-D).
129
Figure 3.11: Expression of p21CIP1/WAF1/Sdi1 in the HMF3AEcoR cells
At 38°C, the thermolabile LT is inactivated, leading to an increase of p53 and therefore the increase of
p21CIP1/WAF1/Sdi1 expression directly downstream (A). If p21 is silenced (A), p21 expression levels stay low
(B) The expression of p21CIP1/WAF1/Sdi1 protein was analyzed by Western blot in cells transfected with
either PLPCX or p21 shRNA at 34 and 38°C (B-D).
130
However, the induction could not be attributed to pRb alone as both E7 and E1A
proteins possess a multifunctional activity able to abrogate other pathways as well as the
pRb pathway.
In addition, Dr. Louise Mansfield had found that p14ARF knockdown was insufficient to
overcome the HMF3A conditional growth arrest. She also tried to knockdown the
p16INK4a but did not manage to get sufficient knockdown of p16INK4A protein levels
despite trying two different shRNA construct namely pRetroSuper-p16#2 utilised by
Wei and colleagues, and pRetroSuper-p16A (Reynolds, Leake et al. 2004).
Ectopic expression of E1A and E7 were tested in both HMF3A EcoR and CL3ECoR cells.
However, because of E7 and E1A proteins multifunctional activity, a more specific
reagent to inactivate the pRb pathway was also investigated; ectopic expression of E2FDB, a mutant of E2F-1shown to be functionally equivalent to the specific inactivation of
the pRb family, was tested.
3.1.5.1 Constitutive Expression of Ad 5 E1A and HPV16
E7
E1A and E7 both possess a similar LxCxE-binding motif to LT (Figure 3.12A, B and C)
which gives them the ability to bind to and inactivate the Rb family members. E1A
retroviral expression construct, pLPC-12SE1AORI (gift from S. Lowe) and an E7
retroviral expression construct, pLPC-E7 (prepared by Dr. Louise Mansfield) were
introduced into the CL3EcoR and HMF3AEcoR cells and assessed for their ability to
complement the conditional growth defect of these cells.
131
Figure 3.12: Conserved Regions of the DNA Tumour Viruses
Diagram of: A: SV40 LT; B: Adenovirus type 5 12S E1A; and C: HPV type 16 E7. NLS: Nuclear
Localisation Signal; CR1: Conserved Region 1; CR2: Conserved Region 2; LXCXE: Rb family binding
motif.
132
10 g of pLPCX-12S E1A, pLPCX and pLPCX-E7 each were packaged, using the 
ecotropic cells. The complementation assay was performed as described previously for
WT LT and the cells were stained with 2% (w/v) methylene blue (Figure 3.7 and 3.8).
In both cases, multiple healthy, growing colonies were observed in the E1A and E7infected CL3EcoR (Figure 3.8, D and F) and HMF3AEcoR (Figure 3.7, D and F) cultures,
but not the control-infected CL3EcoR (Figure 3.8A) and HMF3AEcoR (Figure 3.7A)
cultures. As observed previously while selecting which clone to take forward, CL3 EcoR
cells displayed a higher efficiency to bypass growth arrest with these two constructs than
HMF3AEcoR (Figure 3.7 and 3.8, D and F)
Protein lysates of each condition were analysed by western blot with an E1A or Cyclin
A antibody to assess whether the ectopic expression of pLPC-12S E1A and E7 was
efficient. E1A expression was detected by the E1A antibody (Figure 3.13A) and Cyclin
A antibody should detect if expression of E7 was efficient in inactivating the pRb
pathway
(Figure
3.14A).
Inactivation
of
the
pRb
function,
generally
by
phosphorylation, induces E2F release and the subsequent expression of E2Fdependent proteins, such as CDC2 and Cyclin A (Jarrard, Sarkar et al. 1999). RT-PCR
was performed with E7 specific primers on the RNA from the cells expressing E7. The
results for E1A showed no endogenous expression in the cells with the empty vector
(Figure 3.13B lanes 1 and 2) but a good expression of E1A in the cells with E1A vector
at both 34C and 38C represented by an intense band (Figure 3.13B, lanes 3 and 4).
The Western blot results for E7 showed the endogenous expression of Cyclin A at 34C
and its inactivation at higher temperature (Figure 3.14B, lanes 3 and 4). The RT-PCR
results showed expression of E7 mRNA only in the cells expressing the E7 vector
(Figure 3.14C, lane 3) compared to the controls (Figure 3.14C, lanes 1 and 2).
Constitutive expression of either E1A or E7 was sufficient to overcome the HMF3A
conditional growth arrest at 38C. By extension therefore, abrogation of the pRb
pathway was sufficient to this bypass. However, it is still not possible to attribute this
activity to the specific inactivation of one member of the pRb family as E1A and E7
function does not distinguish between Rb, p107 and p130.
133
Figure 3.13: Expression of pRb in the HMF3AEcoR cells
At 38°C, the thermolabile LT is inactivated, leading to an increase of pRb and therefore the repression of
E2F which fail to activate the transcription of Cyclin A (A). If E1A is expressed, the pRb family is
inactivated and cyclin A get transcribed (A). The expression of E1A protein was analyzed by Western
blot in cells transfected with either PLPCX or plpC-E1A at 34 and 38°C (B).
134
Figure 3.14: Expression of E7 in the HMF3AEcoR cells
At 38°C, the thermolabile LT is inactivated, leading to an increase of pRb and therefore the repression of
E2F which fail to activate the transcription of cyclin A (A). If E7 is expressed, the pRb family is
inactivated and cyclin A get transcribed (A). The expression of cyclin A protein was analyzed by Western
blot (B) and E7 expression was checked by RT-PCR (C) in cells transfected with either PLPCX or plpCE1A at 34 and 38°C.
135
3.1.5.2 Constitutive ectopic expression of E2F-DB mutant
The pRb family controls cell cycle progression by associating with E2Fs which suggest
a great importance of the E2Fs factors in the induction of senescence. More observations
also suggest a possible role for E2F in its regulation: the levels of various E2Fs decrease
during the onset of senescence (Dimri, Hara et al. 1994; Haddad, Xu et al. 1999), and
overexpressed E2F-1 induces both ARF (DeGregori, Leone et al. 1997; Bates, Phillips et
al. 1998; Dimri, Itahana et al. 2000) and premature senescence (Dimri et al, 2000).
In addition, in primary MEFs, the use of a mutant of E2F-1, namely E2F-DB, which
lacks both the C-terminal transactivation and the pRb binding domains but can still bind
to DNA in heterodimeric complex with DP-1, was shown to rescue both p19Arf and
p16INK4a induced growth arrest in mice (Zhang, Postigo et al. 1999). Studies in human
cell systems are limited but one study involving IMR90 human fibroblasts has been
described (Young and Longmore 2004). This paper provided detailed evidence that, like
in the MEFs, E2F-DB can bind to E2F-responsive promoters and displace endogenous
E2Fs in human fibroblasts. Sebastian (Sebastian, Malik et al. 2005) also showed that
expression of E2F-DB was functionally equivalent to pRb family inactivation in MEFs.
Therefore, the hypothesis was that E2F-DB, in the CL3EcoR and HMF3AEcoR cells, would
inactivate the pRb pathway and activate Cyclin A expression (Figure 3.15A).
Therefore, an E2F-DB retroviral expression construct, was constructed by cloning the
E2F-DB open reading frame (gift from Pr. Xin Lu) into the pLPCX vector, and
introduced into the CL3EcoR and HMF3AEcoR cells.
10 g of pLPC-E2F-DB and pLPCX retroviral expression constructs each were
packaged, using the  ecotropic cells. The complementation assay was performed as
described previously for WT LT and the cells were stained with 2% (w/v) methylene
blue (Figure 3.7 and 3.8). In addition, duplicate cultures of HMF3AEcoR expressing
pLPCX-E2F-DB and pLPCX were kept at 34°C for protein and RNA extraction. The
protein lysates were analysed by Western blot with a cyclin A antibody (Figure 3.15B).
136
Figure 3.15: Expression of E2F-DB in the HMF3AEcoR cells
At 38°C, the thermolabile LT is inactivated, leading to an increase of pRb and therefore the repression of
E2F which fail to activate the transcription of cyclin A (A). If E2F-DB is expressed, the pRb pathway is
inactivated and cyclin A get transcribed (A). The expression of cyclin A protein was analyzed by Western
blot (B and E7 expression was checked by RT-PCR (C) in cells transfected with either PLPCX or plpCE1A at 34 and 38°C.
137
The RNA extracts were analysed by RT-PCR with E2F-DB construct specific primers
(Figure 3.15C).
The results of the western blotting anlysis showed endogenous expression of cyclin A
protein in the cells with the empty vector, showing a decrease in cyclin A expression at
38C compared to 34C (Figure 3.15B, lane 1 and 2). The non permissive temperature
leads to LT inactivation, activation of the pRb pathway, binding of the E2F transcription
factor and loss of cyclin A expression. When the cells express E2F-DB, as demonstrated
by the RT-PCR analysis (Figure 3.15C, lane 3 and 4), the cyclin A protein levels are
considerably increased, as E2F is released, especially at 38C where the levels are now
almost equivalent to the ones at 34C (Figure 3.15B, lane 3 and 4).
The E2F-DB infected culture complementation experiment showed a clear rescue by
displaying multiple growing colonies (Figure 3.7 and 3.8, E), but not the controlinfected HMF3A culture (Figure 3.7 and 3.8, A).
Since E2F-DB expression and therefore activation of E2F can both inactivate the pRb
pathway and rescue the cells from senescence, and since E2F is a major downstream
target of pRb, p107 and p130, I can hypothesize that inactivation of the pRb pathway
going through E2F is critical for senescence growth arrest.
3.2 DISCUSSION
Reconstitution of WT LT activity as well as abrogation of either the pRb or the p53
pathway by ectopic expression of WT LT, E1A, E7, E2F-DB, GSE p53 or shRNA
targeting p53 or p21CIP1/WAF1/Sdi1 were shown to be sufficient to complement the
conditional growth arrest at the non permissive temperature in both HMF3AEcoR and
CL3EcoR cells.
This validated the CL3EcoR cells to be representative of the mixed
population to use to dissect the telomere-independent senescence pathways.
138
Abrogation of p53 or p21CIP1/WAF1/Sdi1 permitted bypass of senescence in the all the
clonal cell lines derived from HMF3A cells. In addition, most cells would form growing
colonies after the p53-p21 pathway was abrogated. In comparison, not all clonal cell
lines could be rescued by the abrogation of the pRb pathways and only a minor
component of the HMF3A cells would form growing colonies as a result of the pRb
pathway inactivation.
In addition, it was also concluded that abrogation of either the p53 pathway alone or the
pRb pathway alone, were both sufficient to form growing colonies in both HMF3AEcoR
and CL3EcoR.
3.2.1 Cellular senescence is p53-dependant process in the HMF3A cells
Our findings that abrogation of the p53 pathway alone was sufficient to bypass
senescence in the HMF3A were in agreement with several studies on HDFs indicating a
role of p53 in the induction of senescence (Shay, Pereira-Smith et al. 1991; Brown, Wei
et al. 1997; Wei, Hemmer et al. 2001; Berns, Hijmans et al. 2004). Moreover, loss of
p53 activity alone was reported to be sufficient to impair senescence and promote
tumour progression in an in vivo model of prostate cancer (Chen, Trotman et al. 2005).
Additional data showed that inactivation of the p53 pathway alone was sufficient to
bypass senescence in MEFs (Dirac and Bernards 2003); (Harvey, McArthur et al. 1993)
and HDFs (Wei, Herbig et al. 2003). In the HDFs, not only Rb−/− clones bypassed
senescence but the same phenotype was documented for p21CIP1/WAF1/Sdi1and p53
heterozygous cells, indicating that loss of function of all three genes results in failure to
establish senescence. By contrast, in that study, the abolition of p16INK4A function by the
expression of a p16INK4A -insensitive cyclin-dependent kinase 4 protein or siRNAmediated knockdown provided only minimal lifespan extension that was terminated by
senescence (Wei, Herbig et al. 2003).
139
However, the results described were not in agreement with studies that utilised a TRF2
inhibition model; de Lange and colleagues have used over-expression of a dominantnegative form of TRF2 (TRF2BM) to study a process known as ‗sudden telomere
deprotection‘, where senescence occurs in the absence of telomere shortening
(Karlseder, Broccoli et al. 1999; Stansel, de Lange et al. 2001; Smogorzewska and de
Lange 2002). In addition to showing that WT LT expression was sufficient to bypass
senescence induced in the TRF2 inhibition model, De Lange et al showed that p16INK4a
functioned as a fail-safe mechanism for p53 induced senescence in the absence of a
functional p53 pathway (Smogorzewska and de Lange 2002; Jacobs and de Lange
2004).
This is in agreement with the Campisi group finding that in human fibroblasts and
mammary epithelial cells, expression of telomerase alone does not suffice to reverse
senescence, while p53 inactivation in cells with low levels of p16INK4A (BJ cells)
resumed robust growth. In contrast, cells with high levels of p16INK4A (Wi-38 cells)
failed to proliferate upon p53 inactivation. Therefore, in that case, p16INK4A provided a
dominant second barrier to the unlimited growth of human cells (Beausejour, Krtolica et
al. 2003).
Therefore, the demonstration that abrogation of p53 activity alone was sufficient to
complement the conditional HMF3A growth defect, was contradictory to the Campisi
and de Lange data that implied that both the p53 and pRb pathways must be inactivated.
The fact that endogenous p16INK4a expression was readily detectable in the HMF3A cells
suggests that, in this context, abrogation of senescence by inmactivation of either
pathway happens in spite of p16INK4A expression. The fact that HMF3A cells were
originally derived from adult breast cells may be significant in terms of cell context.
The fact that the TRF2BM inhibition model represents a telomere-dependent system,
unlike the HMF3A system, may help to reconcile the differences observed between
these two systems. Moreover, different methods were used to measure the effects of p53
140
abrogation upon the induction of senescence which may also be an important
consideration.
To validate these findings, both hTERT and the p53 shRNA construct should be
introduced directly into adult mammary fibroblasts and assessed for their ability to
bypass the finite proliferative potential of these cells; it is hypothesised that hTERT and
p53 inactivation will be sufficient to bypass this process, in a similar manner to the
combined activities of hTERT and LT (O'Hare et al, 2001).
As a limitation, however, it is possible that ectopic expression of hTERT and p53 may
not be sufficient to bypass cellular senescence in this context, since there may be
fundamental differences between immortalisation in primary cells, and maintenance of
the immortal state (such as HMF3A cells grown under permissive conditions); for
example, expression of an amino terminal LT mutant that retains p53-binding activity
(dl1135), was sufficient to complement the growth of rat tsa14 cells, but was not able to
immortalise (Powell, Darmon et al. 1999). This indicated that LT functional activities, in
addition to abrogation of p53, were required to initiate this process. It is important to
remember that activities required for initiation may not be required for maintenance.
3.2.2 Senescence is a p21CIP1/WAF1/Sdi1-Dependent Process in the HMF3A Cells
The fact that down-regulation of p21CIP1/WAF1/Sdi1, like p53, by shRNA was sufficient to
bypass the HMF3A conditional growth defect was in accordance other studies of HDFs
(Brown, Wei et al. 1997; Wei, Herbig et al. 2003; Berns, Hijmans et al. 2004).
Moreover, microarray analysis has shown that ectopic expression of p21CIP1/WAF1/Sdi1 in
human fibrosarcoma cells is sufficient to induce changes that are known to occur in
senescent cells, such as the up-regulation of PAI-1 and other extracellular matrix
components and secreted proteases (Chang, Watanabe et al. 2000), whereas downregulation of a large number of genes involved in DNA replication, repair and mitosis by
141
ectopic p21CIP1/WAF1/Sdi1 expression has also been described (Harvat, Wang et al. 1998;
Chang, Watanabe et al. 2000).
In the HMF3A cells cells, it was likely that p21CIP1/WAF1/Sdi1 functioned in a p53dependent process to induce the irreversible growth arrest, since p21CIP1/WAF1/Sdi1 levels
were significantly down-regulated in HMF3AEco R cells complemented for growth at
38°C by the introduction of either p53 shRNA or GSE p53. However, the possibility
that senescence occurred in a p53-independent process could not be excluded. Chen and
colleagues (Chen, Trotman et al. 2005) used two immortalised human models that
lacked functional p53 activity to demonstrate that up-regulation of p21CIP1/WAF1/Sdi1 in
response to Chk2 induction was sufficient to induce senescence.
3.2.3 Inactivation of the pRb pathway in the HMF3A cells
Unlike p53, inactivation of the pRb pathway was technically difficult to achieve in the
HMF3A cells due to the existence of multiple pRb family members and the possibility
that they exhibit functional redundancy. Consequently, a variety of reagents were used
to determine the functional role of this pathway in the induction of the HMF3A growth
arrest.
3.2.3.1 p16INK4a inactivation in the HMF3A Cells
ShRNA targeting has been used previously to impair the negative regulatory activity of
p16INK4a, a CDKI that functions upstream of pRb. Unfortunately, efficient p16INK4a
knockdown could not be obtained by shRNA-targeting in the HMF3A cells despite
trying two different shRNA construct using identical, if not overlapping, target
sequences for p16INK4a that had successfully achieved p16INK4a knockdown in the
litterature (Brookes, Rowe et al. 2002; Narita, Nunez et al. 2003; Wei, Herbig et al.
2003; Berns, Hijmans et al. 2004; Bond, Jones et al. 2004). As a result of efficient
142
p16INK4a knockdown, these studies had concluded that down-regulation of p16INK4a was
not sufficient to prevent the induction of senescence.
The utilisation of alternative, more effective p16INK4a shRNA targets could have
addressed this problem. However, only a limited region of the INK4A locus can be used
to design shRNA constructs to specifically down-regulate p16INK4a, but not p14ARF,
which renders this form of analysis not usable in the HMF3A cells.
3.2.3.2 Bmi-1 Activity in the HMF3A Cells
Bmi-1 ectopic expression was hypothesised to reduce expression from the INK4A locus,
as demonstrated in other HDF strains such as WI-38 HDFs (Itahana, Zou et al. 2003).
However, ectopic Bmi-1 expression had no effect upon p16INK4a expression in the
HMF3A cells.
This supports the hypothesis that human fibroblasts differ in their
sensitivity to Bmi-1, an oncogene that extends the replicative lifespan of fibroblasts by
repressing p16INK4A, apparently because they differ in the level of p16INK4A they express
at senescence (Itahana, Zou et al. 2003; Jacobs and de Lange 2004). This raises the
possibility that human cell strains also differ in the mechanisms that maintain the
senescence state.
3.2.3.3 Ectopic Expression of E1A and E7
Introduction of E1A or E7 into the HMF3A complementation assay was sufficient to
complement the growth of these cells.
These findings are consistent with the
demonstration that E1A expression is sufficient to bypass cellular senescence in primary
IMR-90 HDFs (Serrano, Lin et al. 1997; Narita, Nunez et al. 2003). However, the
efficiency of abrogation by E7 and E1A expression was much lower than abrogation of
the p53 pathway with only a few cells leading to healthily growing colonies, indicating
that the pathways must be parallel and not linear. In addition, expression of E7 and E1A
did not function to reverse senescence in all the clones tested while abrogation of p53 or
143
p21CIP1/WAF1/Sdi1 yielded growing colonies with all clones and the mixed population. This
indicated that the pRb pathway may have a lesser importance in the senescence
mechanism functioning as an alternative secondary pathway to activate the growth
arrest.
This hypothesis is in contradiction with that of Wei and colleagues (Wei, Herbig et al.
2003) who used gene targeting of p53, p21CIP1/WAF1/Sdi1 and pRb, in addition to ectopic
expression of DK and p16INK4A RNAi, to conclude that p53, p21CIP1/WAF1/Sdi1 and pRb
acted in a linear genetic pathway (with pRb acting downstream of p53) to regulate entry
into replicative senescence, and that p16INK4A formed a branch that entered at the level of
pRb (Wei, Herbig et al. 2003). It should be noted that this model is also contradictory
with the senescence induction model proposed by Sharpless and DePinho (Sharpless and
DePinho 2005).
3.2.3.4 Possible Mechanisms by which E1A and E7 Bypass
the Conditional HMF3A Growth Defect
It is possible that both E1A and E7 could have targeted a number of additional cellular
proteins, in addition to pRb and/or p300, to bypass senescence. The mechanism by
which E1A and E7 could have achieved this is unknown, yet the observation that the
HMF3A growth arrest was bypassed by inactivation of p53 and/or p21CIP1/WAF1/Sdi1
suggested that inactivation of the p53 pathway could have been involved.
This
conclusion was in agreement with data derived from the immortalisation of REFs with a
temperature sensitive mutant of p53 (Vousden, Vojtesek et al. 1993); Vousden and
colleagues showed that both E1A and E7 were able to bypass the conditional growth
arrest of these cells and they suggested that E1A and E7 were able to do so by
modulating the activity of p53, without altering its conformation or stability. Quartin
and colleagues also showed that activities mediated by the N-terminal region of LT
could bypass the same conditional growth arrest (Quartin, Cole et al. 1994).
Neither E1A, nor E7 are known to bind directly to p53, yet there is evidence to suggest
that they are both able to inactivate downstream components of this pathway. As an
144
example, E7 has been shown to bind to and inactivate p21CIP1/WAF1/Sdi1 (Helt, Funk et al.
2002). Conversely, there is no evidence of a similar interaction between p21CIP1/WAF1/Sdi1
and E1A. However, p300/CBP, as described above, is a good candidate for this activity
since it is involved in many transcriptional regulation processes by virtue of its
endogenous HAT activity; for example LT, E1A and E7 have all been shown to interact
with p300/CBP, and it has been implicated in the regulation of both p53 phosphorylation
and acetylation status (Pearson, Carbone et al. 2000; Webley, Bond et al. 2000; Pedeux,
Sengupta et al. 2005).
However, the fact that expression of E2F-DB, a repressor of the pRb pathway, is
sufficient alone to bypass senescence in the HMF3A cells indicates that inactivation of
the pRb pathway alone is sufficient to overcome senescence. This does not mean that
there is no interaction between the pRb and the p53 pathways but only that these are not
essential to senescence.
3.2.3.5 p14ARF is not necessary between the p16-pRb and
p21-p53 Pathways
E2F has been shown to directly activate p14ARF in response to various oncogenic stimuli
(DeGregori, Leone et al. 1997; Bates, Phillips et al. 1998; Zhu, DeRyckere et al. 1999;
Parisi, Pollice et al. 2002; Aslanian, Iaquinta et al. 2004). Since p14ARF binds to Mdm2
and impairs the ability of Mdm2 to negatively regulate p53 activity, p14 ARF provides a
link between the pRb and p53 pathways (Dimri et al, 2000). However, evidence from
the HMF3A model indicated that p14ARF did not act upstream of p53 to mediate bypass
of the growth arrest as a functional p14 ARF shRNA construct was insufficient to
complement the growth of these cells as Dr. Louise Mansfield has shown in her thesis.
Therefore, the link between the pRb and p53 pathways via E2F and p14 ARF was
probably not significant in the HMF3A cells. This conclusion is in accordance with the
findings of both Brookes and colleagues (Brookes et al, 2002) and Wei and colleagues
(Wei et al, 2001). Wei and colleagues showed that in HDFs, Ras induced expression of
both p21CIP1/WAF1/Sdi1 and p16INK4A, but not p14ARF; therefore, the induction of
p21CIP1/WAF1/Sdi1 appeared to be p14 ARF-independent. However, these findings were not
145
in agreement with those of Dimri and colleagues (Dimri et al, 2000) who showed that
HDFs deficient in p14 ARF did not undergo senescence. The fact that p14ARF has been
detected at very low levels in normal human cells may have precluded, to some extent,
accurate analysis of p14 ARF activity in these studies.
The HMF3A data was also in contrast to the substantial evidence linking the activity of
p19Arf to the induction of senescence in mice; for example, the p19Arf-p53 pathway has
been shown to play a critical role in the induction of senescence in MEFs (Harvey,
McArthur et al. 1993; Kamijo, Zindy et al. 1997) and a functional screen showed that
down-regulation of p19 Arf was sufficient to rescue premature senescence (Shvarts,
Brummelkamp et al. 2002). Conversely, enforced expression of p19Arf was sufficient to
induce cell cycle arrest in MEFs (Quelle, Zindy et al. 1995). Therefore, the differential
activities of p14ARF in humans and p19Arf in mice may be species- and/or cell typespecific (Brookes, Rowe et al. 2002). General consensus is that ARF is more important
in mice whereas p16INK4A is more important in humans.
It was also possible that, upon the HMF3A temperature shift, E2F activity was able to
induce p53 activation in the absence of p14 ARF induction, similar to the activation of
E2F in response to DNA damage and apoptosis (Tolbert, Lu et al. 2002; Lindstrom and
Wiman 2003). It has been indeed shown that the cyclin A-binding domain of E2F1 can
directly interact with and stabilise p53 in response to DNA damage (Nip, Strom et al.
2001; Hsieh, Yap et al. 2002; Rogoff, Pickering et al. 2002). Such possibilities require
further investigation in the HMF3A cells.
3.2.3.6 E2F-DB bypass the conditional growth arrest by
repressing the pRb pathway
E2F, by its interaction to the pRb protein, form a repressor complex that directly binds to
the DNA of downstream targets. A mutant form of E2F-1, namely E2F-DB, lacks the
pRb-interacting domain as well as the transactivation domain, but is still capable of
binding DNA and displacing endogenous E2F-1/pRb complexes from their binding sites.
146
Sebastian et al (2005) used the E2F-DB construct to mimic pRb family inactivation and
showed that expression of E2F-DB was functionally equivalent to pRb family
inactivation in MEFs.
It has been demonstrated that E2F-DB mutant protein prevents P16INK4A-mediated
growth arrest and allows cells to proliferate at a normal rate, even with a high level of
the P16INK4A inhibitor (Zhang, Postigo et al. 1999). Additional data shows that E2F-DB
can rescue cell cycle arrest induced by ectopic p19Arf expression in MEFs (Rowland,
Denissov et al. 2002). E2F-DB was also able to rescue the proliferative potential of
M33-null MEFs to a normal rate, whereas they were impaired in the progression into the
S phase of the cell cycle in spite of P16INK4a and p19Arf accumulation (Core, Joly et al.
2004). These results are in agreement with the finding that in the HMF3A cells
expression of E2F-DB is sufficient to bypass senescence, suggesting that inactivation of
the pRb alone is sufficient to bypass senescence.
3.2.4 p16-pRb does not always act downstream of p53-p21 to induce
senescence
The development of a conditionally immortal system of human mammary fibroblasts
(HMF3A) cells enabled to define the relative contributions of the p16-pRb and p53-p21
pathways towards senescence, by developing a complementation assay to abrogate each
of these pathways by ectopic expression of various constructs or shRNA mediated
silencing.
Together, these results indicated that in these conditionally immortalised human
mammary fibroblasts, the predominant pathway that induces the irreversible growth
arrest was the p53-p21 pathway since it was most efficiently abrogated when this
pathway was inactivated. Inactivation of the p16-pRb pathway also overcomes the
growth arrest but much less efficiently and in a much smaller number of cells compared
to p53. This indicates that pRb that does not always act downstream of p53-p21 but may
support the idea of parallel pathways.
147
4
ACTIVATION OF THE NF-κB SIGNALLING PROMOTES CELLULAR
SENESCENCE
4.1 SENESCENCE SPECIFIC GENE EXPRESSION RESULTS
4.1.1 Objectives
Genome wide expression profiling technologies have been extensively employed
to identify genes that are differentially expressed in a wide variety of cell types, cancers
and other disease processes. They have also been used to systematically analyse a
variety of cellular processes such as quiescence, stress, replicative and oncogeneinduced senescence and identify the downstream targets of the E2F and p53 pathways.
Previously, cDNA microarrays representing approximately 6000 genes were
used to identify genes that are differentially expressed when the conditionally immortal
mammary fibroblasts undergo irreversible growth arrest, upon activation of the p16-pRb
and p53-p21 pathways. It was discovered that the transcriptional changes that occurred
upon the conditional HMF3A model growth arrest directly correlated with the
transcriptional changes that occurred upon replicative senescence (Hardy, Mansfield et
al. 2005). It appeared that three pathways associated with the induction of replicative
senescence, namely, the p53, pRb and ERK signalling pathways, were also important
regulators of the conditional HMF3A growth arrest. In addition, in silico analysis of the
promoters of genes known to be differential in senescence indicated that NF-κB and CEBP transcription factors may be activated upon senescence.
To investigate further how exactly these pathways affect the genes expression and
furthermore which group of genes preferentially have their expression affected upon
senescence and also to identify novel genes and signaling pathways causal to the
induction of cellular senescence, an extended genome wide microarray expression
profiling analysis (complete coverage of the Human Genome for analysis of over 47,000
148
transcripts) was performed. It was hypothesised that the activity of critical mediators of
senescence would be abrogated in all cells in which the process had been abrogated.
Consequently, I developed a clonal derivative of HMF3A called CL3 EcoR which behave
like HMF3A cells. I first identified changes that were specific for growth arrest by
eliminating changes due to the temperature shift and also identified changes in
expression that could be caused by quiescence.
From there, the study was broadened by overcoming senescence in these cells by
individually abrogating the p53-p21 (with p53 shRNA, p21 shRNA or p53 GSE) and
p16-pRb (with E7, E1A or E2F-DB) pathways as previously described in chapter 1 and
profiling the resultant cells. Upon verification of the expression data by quantitative realtime PCR, the functional activity of the candidate gene was further analysed using the
HMF3A complementation assay to validate the biological effects of the regulated genes .
4.1.2 Why use Microarray Analysis?
The onset of tumourigenesis is a complex mechanism hypothesised to be involving a
limited, but essential set of alterations necessary for tumour development (Hanahan and
Weinberg 2000); namely, self-sufficiency in growth signals, insensitivity to anti-growth
signals, evading of apoptosis, limitless replicative potential, sustained angiogenesis, and
tissue invasion and metastasis (Figure 4.1). The authors state that most, if not all,
cancers will have acquired these capabilities during their development; yet different
genes may be inactivated in different ways, to achieve the same endpoint.
Bypassing senescence represents an example of one of the possible mechanisms utilised
by the cells to acquire limitless replicative potential, making the HMF3A conditional
cells an excellent in vitro model to investigate the signalling pathways that underline
senescence. HMF3A cultures in which senescence has been bypassed by expression of
wt LT, E1A, E7, E2F-DB, p53 shRNA and GSE p53 represented valuable resources
with which to investigate the downstream signalling pathways that induce senescence.
149
Figure 4.1: Cancer: A multistep process
Figure from Hanahan and Weinberg, 2000.
In contrast to normal somatic cells, cancer cells have the potential to proliferate indefinitely and the
acquisition of this limitless replicative potential has been proposed to be one of the six key events required
for malignant transformation .
150
Many experimental techniques were applicable for this investigation, but the rapid
evolution in the development of microarray platform technologies over recent years and
the unbiased nature of the analysis meant that gene expression profiling represented an
attractive approach with which to assess the putative role of many novel genes in this
process. There are many previous examples of the application of this technique in
studies involving senescent HDFs (Cristofalo and Tresini 1998; Shelton, Chang et al.
1999).
A major advantage in applying microarray analysis to the HMF3A system, unlike these
other studies, however, was the rapid and synchronous nature of the conditional growth
arrest.
Additionally, previous study by Hardy et al (2005) has shown that genes
expressed upon HMF3A conditional senescence directly correlated with the
transcriptional changes that occurred upon natural replicative senescence.
4.1.3 Which Microarray Technology?
Advances in sequence selection, sequence clustering, probe modelling, probe
selection, analysis algorithms, and array manufacturing enabled the release of the
Human Genome U133 Set in 2001. In addition, this design incorporated the first
complete draft of the human genome. The GeneChip® Human Genome U133 Plus 2.0
microarray, the latest iteration of the human expression arrays, developed by
Affymetrix, was utilised. Enhancements in array manufacturing, new scanner
technology and improvements in data acquisition allowed better accuracy. The Human
Genome U133 Plus 2.0 Array contained over 54,000 probe sets representing
approximately 38,500 genes on a single array. This increase in feature density allows the
expression of all known transcripts of an organism to be analyzed on a single array. The
sequences from which these probe sets were derived were selected from GenBank®,
dbEST, and RefSeq.
151
4.1.4 Microarray Strategy
It was important to address the issue of variability in the design of the microarray
experiment in order to obtain biologically relevant and reproducible microarray data.
The clonal nature of the CL3EcoR limited, to some extent, the experimental error that
could have occurred as a result of biological variability. However, to further minimise
sources of technical variability, each experimental condition was analyzed using
biological triplicates. Specifically, three cultures were processed in parallel and RNA
was extracted from each culture, as suggested by Lee and colleagues (Lee, Kuo et al.
2000). Assays were performed as systematically as possible to minimise experimental
variation and all samples were processed simultaneously.
It was also important to utilise an appropriate experimental design to ensure the
maximum amount of information was obtained from the microarray data (Larkin, Frank
et al. 2005).
To identify the changes in gene expression that occur upon irreversible growth arrest,
(GA) triplicate independent biological samples of RNA extracted from CL3 EcoR cells
growing at 34C or after growth arrest at 38C for 7 days were analysed by expression
profiling (Figure 4.2). These changes also included gene expression changes that were
non-specific to senescence induction as they could be due to heat shock triggered by the
temperature shift (HS).
To eliminate changes in gene expression due to the temperature shift (HS), two
methods could have been used: Firstly, irreversibly arrested CL3 EcoR cells incubated at
38°C for 7 days can be compared to irreversibly arrested CL3 EcoR cells shifted back
down to 34°C for 7 days. However, there is a possibility that irreversibly arrested
CL3EcoR cells will exhibit an altered pattern of gene expression of heat-shock genes
when compared to proliferating cells. In addition, some genes may even be turned back
on, even though the cells do not divide. Consequently, a second method utilising the
152
Figure 4.2: HMF3AEcoR Microarray strategy
To identify the changes in gene expression that occur upon irreversible growth arrest (GA) and heat shock
(HS) triplicate independent biological samples of RNA extracted from CL3 EcoR (GA) or HMF3S cells
(HS) cells growing at 34C or at 38C for 7 days were analysed by expression profiling.
153
HMF3S cell line was chosen (Figure 4.2).
Triplicate independent RNA samples
extracted from HMF3S cells grown at 34C and after shift up to 38C for 7 days were
analysed. HMF3S cells were immortalised from the same batch of primary human
mammary fibroblasts, using a wild type U19 LT antigen not sensitive to temperature in
conjunction with hTERT, and do not growth arrest upon shift at 38C but continue to
divide and do not express SA-β-Gal.
Each condition was processed in biological
triplicate.
To identify genes that were differential due to the irreversible growth arrest of CL3 EcoR
cells, all changes detected upon shift up of HMF3S cells were eliminated (Figure 4.2).
Genes were considered growth arrest specific when the difference of Log 2 Fold Change
in the gene expression between ―CL3 EcoR 38 versus 34‖ and ―HMF3S 38 versus 34‖ was
>1 or < -1 (equivalent to a 2 fold up- or down-regulation).
In order to identify genes whose expression may also be altered by serum starvation
resulting in quiescence, a state of reversible growth arrest, CL3 EcoR cells were serum
starved for 7 days at 34C, triplicate independent RNA samples extracted and compared
to profiles of CL3EcoR cells cultured at 34C; shown schematically in Figure 4.3.A.
If changes in gene expression are specific for the senescence growth arrest, they should
be reversed upon its abrogation. To identify if changes in gene expression would be
reversed, triplicate independent cultures of CL3EcoR cells after complementation of the
growth defect with SV40 LT antigen, Ad5 E1A 12S, HPV16 E7, E2F-DB, p53GSE,
p53shRNA and p21shRNA were derived and profiled. The data for each rescued culture
was averaged, compared to its appropriate control cells to obtain the set of differential
genes which were then compared to the differential GA data set (shown schematically in
Figure 4.3B).
154
Figure 4.3: Microarray strategy for the complementations
To identify the changes in gene expression that occur upon quiescence (Q)
triplicate independent
biological samples of RNA extracted from CL3EcoR cells growing at 34C in a normal media or in a FCS
depleted media (quiescence) for 7 days were analysed by expression profiling (A). To identify if changes
in gene expression would be reversed, triplicate independent cultures of CL3 EcoR cells after
complementation of the growth defect with SV40 LT antigen, Ad5 E1A 12S, HPV16 E7, E2F-DB,
p53GSE, p53shRNA and p21shRNA were derived and profiled (B).
155
4.1.5 Microarray procedure
To perform the microarray procedure, total RNA was extracted from CL3 EcoR cells
incubated at 34 and 38C to prepare the reference RNA samples (Figure 4.2 and 4.3A
and B) or at 38C with the various constructs described in chapter one (PLPCX, PLPC
E7, PLPC E1A, PLPC E2F-DB, PLXIP GSE p53, pRS Lamin A/C shRNA, pRS p21
shRNA and pRS p53 shRNA) to prepare the different conditions samples to analyse.
Additional total RNA was extracted from quiescent CL3 EcoR cells and HMF3S cells to
prepare the heat shock and quiescence samples. RNA was extracted from biological
triplicate cultures using Trizol (Invitrogen), frozen and sent for analysis at the Memorial
Sloan Kettering Cancer Center Microarray facility.
4.1.6 Microarray results
Application of the strategy outlined in Figure 4.2 identified 3059 up-regulated
transcripts of which 816 were up-regulated >2 fold and 5005 were down-regulated, 961
of which were down-regulated >2 fold. The top 24 up- and down-regulated transcripts
ranked according to log2 Fold Change are shown in Table 4.1 A and B; the complete
lists are in Supplementary Tables S4.1 and S4.2 (supplementary on a CD). Three of the
top four most highly down-regulated transcripts (NUF2, SLC25 and NDC80) are all
components of the NDC80 kinetochore complex; NUF2 was decreased 23 fold (P-value
1.88E-23), SPC25 18 fold (P-value 1.39E-24) and NDC80 17 fold (P-value 1.12E-21)
respectively. All of the top down-regulated transcripts yielded highly significant pvalues. Four of the top five most highly up-regulated transcripts correspond to the same
gene, CLCA family member 2, chloride channel regulator. This was due to the gene
being present in four different locations on the chips. It also validates the accuracy of the
microarray if the same gene represented by different oligos gives similar results. The
fold increase in expression of CLCA2 was 28 (P-value 1.87E-12), 23 (P-value 5.62E13), 19 (P-value 6.25E-12) and 15 (P-value 1.85E-12) respectively.
156
A
logFC
GA
P.val
logFC
HS
P.Val
logFC Q
P.Val
CLCA family member 2, chloride channel regulator
4.79
1.9E-12
0.18
8.0E-01
0.89
8.1E-02
CLCA family member 2, chloride channel regulator
4.51
5.6E-13
-0.08
9.0E-01
0.67
1.5E-01
interleukin 33
4.24
8.0E-10
2.58
4.1E-05
3.06
9.5E-07
Probe
Symbol
Description
217528_at
CLCA2
206165_s_at
CLCA2
209821_at
IL33
206166_s_at
CLCA2
CLCA family member 2, chloride channel regulator
4.22
6.3E-12
-0.16
8.1E-01
0.09
8.7E-01
206164_at
CLCA family member 2, chloride channel regulator
3.87
1.9E-12
-0.10
8.6E-01
0.57
1.7E-01
243036_at
CLCA2
RP4692D3.1
hypothetical protein LOC728621
3.75
3.5E-12
-0.18
7.4E-01
1.55
2.0E-04
225895_at
SYNPO2
synaptopodin 2
3.74
3.0E-10
0.45
4.5E-01
-1.62
9.3E-04
glutaminase
3.53
8.5E-16
-0.62
4.8E-02
1.35
7.9E-06
ABI gene family, member 3 (NESH) binding protein
3.52
1.0E-13
0.07
8.8E-01
3.95
4.3E-15
butyrylcholinesterase
similar to ankyrin repeat domain 20 family,
member A1
3.51
2.9E-07
1.65
1.3E-02
0.54
4.1E-01
3.51
1.5E-08
-0.46
4.9E-01
-0.34
5.6E-01
203158_s_at
GLS
220518_at
ABI3BP
205433_at
BCHE
237737_at
LOC727770
223734_at
OSAP
ovary-specific acidic protein
3.44
1.8E-14
0.96
3.7E-03
-0.76
1.4E-02
201860_s_at
PLAT
plasminogen activator, tissue
3.42
1.2E-17
1.71
9.9E-09
0.45
6.2E-02
210118_s_at
IL1A
3.42
1.7E-10
0.22
7.0E-01
3.82
1.1E-11
226757_at
IFIT2
interleukin 1, alpha
interferon-induced protein with tetratricopeptide
repeats 2
3.39
8.7E-08
0.34
6.4E-01
1.62
4.3E-03
cadherin 10, type 2 (T2-cadherin)
3.37
9.1E-12
1.77
2.5E-05
-1.77
9.8E-06
interleukin 1, beta
3.33
1.9E-08
0.06
9.4E-01
3.54
5.2E-09
spermatogenesis associated 18 homolog (rat)
3.31
1.3E-15
0.09
8.2E-01
-0.38
1.6E-01
3.29
1.0E-16
0.18
5.7E-01
1.13
1.5E-05
interleukin 1, beta
pregnancy-associated plasma protein A,
pappalysin 1
3.29
7.2E-09
0.38
5.4E-01
3.43
3.2E-09
3.25
2.2E-11
1.69
4.8E-05
2.77
1.1E-09
glutaminase
3.23
4.1E-18
-0.65
5.1E-03
1.42
1.6E-08
220115_s_at
CDH10
205067_at
IL1B
229331_at
SPATA18
238733_at
39402_at
228128_x_at
203159_at
IL1B
PAPPA
GLS
223395_at
ABI3BP
ABI gene family, member 3 (NESH) binding protein
3.2
1.6E-09
0.87
7.8E-02
3.39
4.2E-10
225720_at
SYNPO2
synaptopodin 2
3.16
1.5E-15
-0.05
9.0E-01
-1.02
1.3E-04
157
B
logFC
GA
P.val
logFC
HS
-4.52
1.9E-23
0.46
4.1E-02
-0.40
5.4E-02
discs, large (Drosophila) homolog-associated protein 5
SPC25, NDC80 kinetochore complex component, homolog
(S. cerevisiae)
NDC80 homolog, kinetochore complex component (S.
cerevisiae)
-4.21
3.0E-26
0.28
8.3E-02
-0.72
4.4E-06
-4.20
1.4E-24
-0.13
5.5E-01
-0.37
3.7E-02
-4.08
1.1E-21
0.43
6.9E-02
-0.05
8.4E-01
cell division cycle 20 homolog (S. cerevisiae)
asp (abnormal spindle) homolog, microcephaly associated
(Drosophila)
-4.06
1.3E-25
-0.06
7.8E-01
-1.34
1.1E-11
-3.99
3.0E-22
0.52
1.5E-02
-0.17
4.2E-01
NIMA (never in mitosis gene a)-related kinase 2
asp (abnormal spindle) homolog, microcephaly associated
(Drosophila)
-3.97
1.3E-25
0.24
1.5E-01
-0.43
4.2E-03
-3.97
2.1E-22
0.84
7.7E-05
-0.11
5.9E-01
232278_s_at DEPDC1
DEP domain containing 1
-3.77
2.3E-22
0.90
1.5E-05
-0.28
1.4E-01
219148_at
PDZ binding kinase
-3.71
5.3E-26
-0.10
5.6E-01
-0.50
2.3E-04
Probe
Symbol
223381_at
NUF2
203764_at
DLGAP5
209891_at
SPC25
204162_at
NDC80
202870_s_at
CDC20
219918_s_at
ASPM
204641_at
NEK2
232238_at
ASPM
PBK
Description
NUF2, NDC80 kinetochore complex component, homolog
(S. cerevisiae)
P.Val logFC Q P.Val
215942_s_at
GTSE1
G-2 and S-phase expressed 1
-3.62
3.9E-19
0.38
1.4E-01
-0.15
5.5E-01
222039_at
KIF18B
kinesin family member 18B
-3.57
2.8E-22
-0.01
9.8E-01
-0.38
3.1E-02
228323_at
CASC5
cancer susceptibility candidate 5
-3.56
9.9E-23
0.04
8.8E-01
-0.36
3.5E-02
204962_s_at
CENPA
centromere protein A
-3.52
6.3E-25
0.46
2.5E-03
-0.06
6.9E-01
201291_s_at
1552619_a_a
t
TOP2A
topoisomerase (DNA) II alpha 170kDa
-3.44
1.4E-24
0.51
9.1E-04
0.02
9.1E-01
ANLN
anillin, actin binding protein
-3.42
7.2E-25
-0.20
2.0E-01
-1.01
1.4E-09
207165_at
HMMR
hyaluronan-mediated motility receptor (RHAMM)
-3.39
1.1E-23
0.71
2.0E-05
-0.06
7.2E-01
218755_at
KIF20A
kinesin family member 20A
-3.39
1.1E-22
0.66
1.9E-04
-0.98
9.6E-08
236641_at
KIF14
-3.37
2.5E-23
0.71
2.6E-05
-0.04
8.3E-01
203755_at
BUB1B
kinesin family member 14
BUB1 budding uninhibited by benzimidazoles 1 homolog
beta (yeast)
-3.36
3.1E-25
0.18
2.5E-01
-0.30
2.2E-02
204318_s_at
GTSE1
G-2 and S-phase expressed 1
-3.36
1.1E-22
0.36
4.1E-02
-0.29
7.5E-02
202240_at
PLK1
polo-like kinase 1 (Drosophila)
-3.35
1.5E-26
0.14
3.1E-01
-1.04
3.7E-12
204444_at
KIF11
kinesin family member 11
-3.34
1.3E-23
-0.08
6.7E-01
-0.50
8.5E-04
218009_s_at
PRC1
protein regulator of cytokinesis 1
-3.33
2.4E-25
0.24
9.0E-02
-0.59
1.2E-05
Table 4.1: Senescence specific changes in gene expression
Log2 fold changes in gene expression that occur upon irreversible growth arrest are indicated as GA, upon
shift up of HMF3S cells from 34C to 38C are indicated as HS and upon serum starvation are indicated
as Q. Up-regulated transcripts are indicated in green whereas down-regulated transcripts are in red.
Results for the top 24 up- (A) and down-regulated (B) transcripts upon growth arrest, heat shock and
quiescence are shown.
158
The P-values for the up-regulated transcripts were lower than those for the downregulated transcripts but were still highly significant and less than E-07.
To refine the differential gene expression data set, several comparisons were carried out
including identifying genes associated with quiescence (Figure 4.3A). This control is
designed to identify whether gene expression is also altered upon quiescence, namely,
serum-starvation.
The results are also presented in Table 4.1 A and B as log2 FC Q for the top 24 up- and
down-regulated transcripts. Interestingly, many of the top 24 up-regulated genes were
also highly up-regulated upon serum starvation (Table 4.1A); for example IL33, ABI3P,
IL1A, IL1B, and PAPPA.
The results obtained after rescue with the various constructs for the top 24 up- and
down-regulated genes upon irreversible growth are shown in Table 2A&B; the complete
data sets are in Supplementary Tables S4.1 and S4.2 (on CD).
The results showed that when growth arrest was overcome, differential expression was
reversed; down-regulated genes (red) were up-regulated (green) whereas up-regulated
genes (green) were suppressed (red). The global reversion, for the quasi-totality of the
genes, upon complementation by abrogation of the pRb pathway or the p53 pathway, is
impressive and further reinforces the involvement of these genes in the senescence
mechanisms. However, the fold change was not always the same across the different
complementations eg. for CLCA2 (Table 4.2A), the fold suppression upon rescue with
Ad5 E1A 12S, p53shRNA and GSE53 was about 30 fold whereas with HPV16 E7, E2FDB and p21shRNA, the fold change was 4 fold. Although these differences may be due
to level of expression of the complementing gene, they are more likely to reflect the
rescuing pathways as illustrated by the changes in expression of MDM2 (HMD2, Table
4.2C) which is an E3 ubiquitin ligase that associates with p53 and maintains it at a low
level; up-regulation of p53 results in up-regulation of MDM2. When CL3EcoR cells
undergo growth arrest MDM2 was up-regulated; expression of all three MDM2 features
was increased.
159
A
Symbol
CLCA2
CLCA2
IL33
CLCA2
CLCA2
Description
CLCA family member 2, chloride channel
regulator
CLCA family member 2, chloride channel
regulator
interleukin 33
CLCA family member 2, chloride channel
regulator
CLCA family member 2, chloride channel
regulator
logFC
GA
logFC logFC logFC
log FC
logFC
logFC
GSE_p5 pRS_p5 pRS_p2
logFC Q
HS
wt_LT
E1A
3
3
1
logFC logFC
E7
E2F-DB
4.79
0.18
0.89
-4.70
-5.46
-6.13
-2.51
-5.59
-1.97
-1.83
4.51
-0.08
0.67
-4.45
-5.26
-5.94
-2.49
-5.08
-1.76
-1.86
4.24
2.58
3.06
-6.75
-5.33
-2.33
-2.83
-7.10
-4.14
-3.66
4.22
-0.16
0.09
-3.87
-4.48
-5.21
-2.29
-4.46
-1.60
-1.89
3.87
-0.10
0.57
-3.18
-3.37
-4.41
-2.58
-3.45
-1.61
-1.52
RP4692D3.1
hypothetical protein LOC728621
3.75
-0.18
1.55
-4.26
-2.96
-2.19
-2.63
-4.59
-2.85
-2.98
SYNPO2
synaptopodin 2
3.74
0.45
-1.62
-1.39
-0.55
-2.68
-2.73
-3.41
-2.20
-3.13
GLS
glutaminase
3.53
-0.62
1.35
-1.84
-0.78
-0.82
-1.23
-1.82
-1.64
-1.24
ABI3BP
ABI gene family, member 3 (NESH)
binding protein
3.52
0.07
3.95
-2.32
-1.84
-1.64
-1.29
-2.75
-1.13
-0.88
BCHE
butyrylcholinesterase
3.51
1.65
0.54
-1.55
-1.73
-1.60
-1.67
0.48
-1.55
-0.36
LOC727
770
similar to ankyrin repeat domain 20 family,
member A1
3.51
-0.46
-0.34
-4.08
-4.32
-3.71
1.17
-2.83
-1.35
0.01
OSAP
ovary-specific acidic protein
3.44
0.96
-0.76
-0.50
-0.68
-1.12
-1.60
0.48
-0.76
-0.11
PLAT
plasminogen activator, tissue
3.42
1.71
0.45
-1.54
-1.12
-1.10
-1.92
-2.69
-1.57
-1.67
IL1A
interleukin 1, alpha
3.42
0.22
3.82
-4.36
-3.28
-0.70
-0.26
-5.18
-1.05
-0.04
IFIT2
interferon-induced protein with
tetratricopeptide repeats 2
3.39
0.34
1.62
-0.70
-1.67
-0.53
-1.42
-3.47
-0.86
-2.46
CDH10
cadherin 10, type 2 (T2-cadherin)
3.37
1.77
-1.77
-2.78
-2.39
-3.19
-0.50
-0.97
-1.24
-1.65
IL1B
interleukin 1, beta
3.33
0.06
3.54
-4.60
-4.13
-1.41
-1.48
-5.44
-1.98
-0.86
SPATA1
8
238733
_at
spermatogenesis associated 18 homolog
(rat)
3.31
0.09
-0.38
-1.78
-3.10
-2.79
-0.19
-0.28
-0.91
-1.63
3.29
0.18
1.13
-2.10
-2.94
-3.22
0.47
2.08
-0.35
0.14
IL1B
interleukin 1, beta
3.29
0.38
3.43
-4.58
-3.83
-1.31
-1.37
-5.89
-1.82
-0.80
PAPPA
pregnancy-associated plasma protein A,
pappalysin 1
3.25
1.69
2.77
-3.90
-1.92
-1.80
-0.16
-5.56
-1.76
-1.35
GLS
glutaminase
3.23
-0.65
1.42
-1.60
-0.26
-1.01
-1.18
-1.68
-1.63
-1.26
ABI3BP
ABI gene family, member 3 (NESH)
binding protein
3.2
0.87
3.39
-3.68
-2.11
-1.16
-1.20
-7.66
-1.70
-1.52
SYNPO2
synaptopodin 2
3.16
-0.05
-1.02
-0.97
-0.73
-2.04
-2.08
-1.67
-1.16
-1.59
160
B
Symbol
NUF2
DLGAP5
Description
logFC logFC logFC
logFC GSE_p pRS_p5 pRS_p2 logFC
wt_LT
E7
53
3
1
logFC
GA
logFC
HS
logFC
Q
-4.52
0.46
-0.40
4.60
4.15
3.77
3.21
3.71
4.79
3.93
-4.21
0.28
-0.72
4.11
3.54
3.41
3.19
3.36
4.30
3.66
-4.20
-0.13
-0.37
3.43
3.22
3.21
2.93
2.96
3.63
3.14
NUF2, NDC80 kinetochore complex component,
homolog (S. cerevisiae)
discs, large (Drosophila) homolog-associated
protein 5
logFC logFC
E1A E2F-DB
NDC80
SPC25, NDC80 kinetochore complex component,
homolog (S. cerevisiae)
NDC80 homolog, kinetochore complex
component (S. cerevisiae)
-4.08
0.43
-0.05
3.97
4.29
3.50
2.95
3.13
3.39
2.98
CDC20
cell division cycle 20 homolog (S. cerevisiae)
-4.06
-0.06
-1.34
3.73
3.41
3.54
3.03
3.38
3.53
3.59
ASPM
asp (abnormal spindle) homolog, microcephaly
associated (Drosophila)
-3.99
0.52
-0.17
4.66
4.19
3.93
3.40
3.94
5.11
4.29
NEK2
NIMA (never in mitosis gene a)-related kinase 2
-3.97
0.24
-0.43
3.99
3.57
3.22
3.06
3.42
3.99
3.71
ASPM
asp (abnormal spindle) homolog, microcephaly
associated (Drosophila)
-3.97
0.84
-0.11
4.30
3.91
3.62
3.09
3.41
4.81
3.86
DEPDC1
DEP domain containing 1
-3.77
0.90
-0.28
3.99
3.73
3.53
3.06
3.41
3.89
3.82
PBK
PDZ binding kinase
-3.71
-0.10
-0.50
3.34
3.45
2.93
2.81
3.02
3.27
3.17
GTSE1
G-2 and S-phase expressed 1
-3.62
0.38
-0.15
3.16
2.68
2.39
2.15
2.37
3.11
2.39
KIF18B
kinesin family member 18B
-3.57
-0.01
-0.38
3.35
3.47
3.09
2.69
2.83
3.18
2.90
CASC5
cancer susceptibility candidate 5
-3.56
0.04
-0.36
3.62
3.30
3.08
2.91
2.76
3.50
3.09
CENPA
centromere protein A
-3.52
0.46
-0.06
3.56
3.35
2.98
2.44
2.76
3.80
2.89
TOP2A
topoisomerase (DNA) II alpha 170kDa
-3.44
0.51
0.02
3.39
3.51
2.80
2.56
2.81
3.52
2.94
ANLN
anillin, actin binding protein
-3.42
-0.20
-1.01
3.54
3.34
3.08
2.69
2.95
3.15
3.32
HMMR
hyaluronan-mediated motility receptor (RHAMM)
-3.39
0.71
-0.06
3.17
3.04
3.08
2.88
3.06
2.78
3.21
KIF20A
kinesin family member 20A
-3.39
0.66
-0.98
3.78
3.37
2.89
2.78
3.11
3.64
3.06
KIF14
kinesin family member 14
-3.37
0.71
-0.04
3.62
3.03
3.13
2.70
2.80
3.91
3.13
BUB1B
BUB1 budding uninhibited by benzimidazoles 1
homolog beta (yeast)
-3.36
0.18
-0.30
3.21
3.03
2.81
2.41
2.60
3.22
2.72
GTSE1
G-2 and S-phase expressed 1
-3.36
0.36
-0.29
3.30
2.87
2.36
2.17
2.53
3.36
2.61
PLK1
polo-like kinase 1 (Drosophila)
-3.35
0.14
-1.04
2.96
2.24
2.61
2.21
2.46
2.61
2.58
KIF11
kinesin family member 11
-3.34
-0.08
-0.50
3.19
2.87
2.37
2.07
2.46
3.14
2.43
PRC1
protein regulator of cytokinesis 1
-3.33
0.24
-0.59
3.14
2.81
2.58
2.25
2.66
3.33
2.69
SPC25
C
Log FC GA P. Val
Q
P. Val
HS
GSE
P53
P21
P. Val wt LT P. Val p53 P. Val RNAi P. Val RNAi P. Val E1A P. Val
E2F
E7 P. Val DB P. Val
MDM2 1.91 7.E-10 0.28 3.E-01 -0.08 8.E-01 -1.60 5.E-08 -2.78 7.E-14 -2.41 7.E-12 0.79 4.E-03 1.49 8.E-08 0.11 8.E-01 0.73 6.E-03
MDM2 1.70 4.E-11 -0.10 6.E-01 -0.10 7.E-01 -1.27 4.E-08
-2.42
5.E-15 -2.20 1.E-13 0.73 9.E-04 1.59 8.E-11 0.24 3.E-01 0.60 4.E-03
MDM2 1.42 8.E-09 -0.26 2.E-01 0.33 2.E-01 -1.14 9.E-07 -1.68 2.E-10 -1.73 2.E-10 0.47 4.E-02 1.39 5.E-09 0.09 8.E-01 0.52 2.E-02
Table 4.2: Senescence specific changes in gene expression with complementation
Log2 fold changes in gene expression that occur upon irreversible growth (GA), upon shift up of HMF3S
cells from 34C to 38C (HS) and upon serum starvation (Q). If changes in gene expression are specific
for the senescence growth arrest, they should be reversed upon its abrogation. Up-regulated transcripts are
indicated in green whereas down-regulated transcripts are in red. Results for the top 24 up- (A) and downregulated (B) transcripts after complementation with the indicated constructs as well as the changes in
expression of the three MDM2/HDM2 features (C) are shown.
161
When growth arrest was abrogated using WT LT antigen, p53GSE or p53shRNA that
directly inhibit p53 activity, MDM2 expression was reversed for all three features.
However when growth arrest was abrogated with Ad5 E1A 12S, HPV16 E7 and E2FDB or p21shRNA, none of which are known to directly act on p53, MDM2 expression
levels remained up-regulated, although the fold up-regulation was reduced by HPV16
E7 and E2F-DB or upon silencing of p21CIP1.
4.1.7 Validation of Microarray Data
4.1.7.1 Why Validate?
The fact that considerable inconsistencies have been observed between different
microarray studies in terms of the different platform technologies, methodologies,
protocols and analyses (Bammler, Beyer et al. 2005; Irizarry, Warren et al. 2005; Larkin,
Frank et al. 2005), emphasised the need to independently verify the microarray data. A
number of techniques were available to facilitate this, including semi-quantitative RTPCR, real-time PCR, northern blot and ribonuclease protection assay. However, due to
the number of genes to verify and the short time scale left to do so, I collaborated with
Biotrove Inc. to use their signal transduction panel of ~600 genes for real time PCR
analysis. I provided Biotrove with the mRNA, Dr Elen Ortenberg did the Q-PCR
analysis and the initial data analysis.
4.1.7.2 Real time validation of expression data using the
BioTrove Open Arrays
To confirm the expression profiling data, the OpenArray Pathways Human Signal
Transduction Panel Analysis developed by Biotrove was utilised.
Duplicate RNA
samples that had been used for the profiling studies extracted from CL3 EcoR cells grown
at 34c and after shift to 38C for 7 days or HMF3S cells grown at 34c and 7 days after
shift to 38C were used. The 630 genes comprising the Signal Transduction panel are
presented in Supplementary Table S4.3 (supplementary on a CD). Ct values were
162
normalised using the geomean of 18 housekeeping genes. Cts were calculated for each
sample and averaged for each group. The biological replicates were highly reproducible
(Figure 4.4A); the average standard deviation in Ct between biological replicates was
<0.3Ct. First, filtering of growth arrest genes whose expression was very low (Ct >22)
and genes whose Ct confidence<300 was applied. In addition, fluidics filters were
applied too with ROX>1500 and SYBR> 400. The differences in the Ct (Ct) were
then calculated for the 318 remaining genes and used to determine the fold change in
expression (FC = 2Ct).
The results of the Biotrove Real Time qPCR Openarray were plotted against the
Affymetrix array data for both growth arrest and heat shock (Figure 4.4B). Comparison
of the log2 fold changes in expression show a plot concentrated around a linear
regression trend line with a coefficient of correlation of respectively 0.62 and 0.59 for
the totality of the points. This correlation coefficient was not entirely satisfactory in
terms of statistics to confidently validate the correlation between the two techniques.
However, both techniques are known to be very variable and therefore the threshold of
the correlation coefficient to be expected for confirmation of the method has to be
lowered. In addition, if you remove only 4 points (out of 318) that seem to be aberrant
results on the graph, the coefficients of correlation climb to 0.80 for the growth arrest
and 0.72 for the heat shock, which validates more than satisfactorily the microarray
results. It could be that these 4 points are false as correct clone annotation was not
assessed for each of the differentially expressed genes against Basic Local Alignment
Search Tool (BLAST) so correct clone sequence for each spot cannot be guaranteed.
Another possibility could be that the microarray chip itself and the individual spots on it
could have been subject to cross contamination.
163
Figure 4.4: Validation of microarray data by real-time qPCR
(A) Reproducibility between samples: RNA samples used for expression profiling were analysed by real
time qPCR using the OpenArrayTM Pathways Human Signal Transduction panel. Ct values were
normalised using the geomean of 18 house keeping genes. ΔCts were calculated for each sample and
averaged for each group. The graphs show that the duplicate RNA samples were highly reproducible. (B)
Comparison of Affymetrix with OpenArrayTM: Comparison of the log2fold changes in expression
obtained using the Affymetrix U133 Plus2 chips versus real time qPCR for growth arrest (CL3 EcoR at 34°C
and 38°C) and heat shock (HMF3S at 34°C and 38°C). Genes whose expression level was very low
(Ct>22) and whose Ct confidence was <300 were not considered. The fluidics filters ROX>1500 and
SYBR>400 were also applied.
164
Nearly 80% of the genes show concordant direction changes in expression between the
two methods, for both growth arrest (79.9%) and heat shock (78.6%) (Table 4.3A),
which confirms the microarray data. However, the actual values of the log-fold changes
detected by real time qPCR or by hybridisation analysis were different (full results of the
Biotrove experiment for both HMF3S and CL3ECoR cells are presented in Supplementary
Table S4.4 and S4.5 (Supplementary on a CD). When looking only at the genes that are
up-regulated on the Affymetrix chip (Table 4.3B); the results are much better with a
concordance of 93.7% and 90.3% respectively for growth arrest and heat shock.
Similarly, when looking at the genes down-regulated on the Affymetrix chip (Table 4.
3C), the results are not as good with a concordance of 66.0% and 67.7% only
respectively for growth arrest and heat shock. This would mean that up-regulated results
are more trustworthy than the down-regulated ones and that there is a bias in one or both
of the methods when genes are down-regulated.
4.1.8 Comparison of genes differentially expressed upon senescence with the
meta-signature of genes over-expressed in cancer
To determine if any of the senescence growth arrest genes have previously been
identified to be important for cancer development, the growth arrest data set was
overlapped with the meta-signatures of genes over-expressed upon neoplastic
transformation and in undifferentiated cancer (Rhodes, Yu et al. 2004). The neoplastic
transformation meta-signature comprises 67 over-expressed genes presented in
Supplementary Table S4.6 (on a CD); 33 of these were found to be down-regulated and
10 were up-regulated upon growth arrest in CL3 EcoR cells (Table 4.4 and Supplementary
Table S 4.6). The meta-signature of genes over-expressed in undifferentiated cancer
comprises 69 genes of which 5 were up-regulated and 46 were down-regulated upon
growth arrest in CL3EcoR cells (Table 4.4 and Supplementary Table S4.6). This indicated
that 49% and 67% of the genes over-expressed upon neoplastic transformation and in
undifferentiated cancer were also down-regulated upon senescence.
165
Table 4.3: Results of comparison Affymetrix with OpenArrayTM
(A)Concordance total results: Numbers of concordant direction changes in expression between the two
methods, for both growth arrest and heat shock. (B)Concordance results of up-regulated genes upon
senescence by Affymetrix: Numbers of concordant direction changes in expression between the two
methods, for both growth arrest and heat shock. (C)Concordance results of down-regulated genes upon
senescence by Affymetrix: Numbers of concordant direction changes in expression between the two
methods, for both growth arrest and heat shock .
Table 4.4: Metasignatures of neoplastic transformation and undifferentiated cancer
The growth arrest differential data set was overlapped with the meta-signatures of genes over-expressed
upon neoplastic transformation and undifferentiated cancer (Rhodes et al 2004). The results of the overlap
are shown both in number of genes and percentage of the the total number of genes studied in the Rhodes
paper.
166
This further validates a definite correlation between cancer expression changes and
senescence expression changes making these results interesting to study further to
understand the importance of senescence in the cancer mechanisms and the roles played
by these changing targets. I believe this is the first time an overlap was made between
senescence and cancer expression data. The metasignatures of genes that are upregulated upon neoplastic transformation or undifferentiated cancer show that nearly
50% of these genes were down-regulated upon senescence which highlights the
importance of this barrier to cancer development.
4.2 BIOLOGICAL VALIDATION BY LENTIVIRAL SILENCING OR
ECTOPIC EXPRESSION
4.2.1 Objectives
To biologically validate the results of the microarray and to check whether up-and
down-regulation of genes were causal to senescence or merely a consequence of it, in
vitro validation was designed.
Genes down-regulated upon senescence can be tested by ectopic expression to define
whether this down-regulation was essential to senescence. Up-regulated genes can be
tested by silencing expression. The silencing strategy chosen to validate the micro-array
results was Lentiviral shRNA silencing using pGIPZ lentiviral shmiRs from Open
Biosystems.
4.2.2 Up-regulated genes upon senescence: Does Silencing bypass the growth
arrest?
Genes up-regulated upon senescence that possessed some link to the cell cycle or to
cancer in the literature were chosen. Silencing of these genes were performed either by
167
multiple individual constructs or by a mix of several constructs and were tested by
complementation assay in the CL3 ECoR cells to assess whether their silencing would
bypass the growth arrest and therefore would place them as key effectors in the
senescence pathways.
4.2.2.1 CLCA2 silencing bypassed senescence at a low
level
CLCA2 belongs to the calcium sensitive chloride conductance protein family. It is
expressed predominantly in trachea and lung and suggested to play a role in the complex
pathogenesis of cystic fibrosis. It may also serve as an adhesion molecule for lung
metastatic cancer cells, mediating vascular arrest and colonization, and furthermore, it
has been implicated to act as a tumour suppressor gene for breast cancer. CLCA2 was
the most up-regulated target upon senescence with 4 oligos coming up in the top 5 upregulated genes.
In addition, all four members of the human CLCA gene family cluster on the short arm
of chromosome 1 at 1p31, a region that is frequently deleted in breast cancer (Hoggard,
Brintnell et al. 1995; Nagai, Negrini et al. 1995; Tsukamoto, Ito et al. 1998; Su, Roberts
et al. 1999; Sossey-Alaoui, Kitamura et al. 2001). However, only CLCA2 gene
expression was shown to be down-regulated in breast cancer and was suggested to act as
a tumour suppressor (Gruber and Pauli 1999; Li, Cowell et al. 2004). Interestingly, Elble
and colleagues have shown that acute expression of CLCA2 induces a senescence like
growth arrest (Walia, Ding et al. 2009).
For these reasons, it was interesting to see whether the level expression of CLCA2 and
its potential tumour suppressor activity would have an effect on senescence in the
CL3EcoR cells.
A complementation assay was performed in CL3EcoR cells with a mix of 3 lentiviral
GIPZ
CLCA2
silencing
constructs
namely:
168
human GIPZ
lentiviral
shMiR
V2LHS_197853, human GIPZ lentiviral shMiR V2LHS_197750 and human GIPZ
lentiviral shMiR V2LHS_199854.
The cells were stained after 3 weeks at 38°C. The results show in Figure 4.5 that CLCA2
silencing with a mix of 3 constructs permits rescue, at a low level, when compared to the
negative control. A repeat of this experiment (data not shown) showed an even lower
level of rescue.
4.2.2.2 AK3L1 silencing bypassed senescence
AK3L1 has been described as a gene over-expressed in fibroblasts undergoing
replicative senescence (Binet, Ythier et al. 2009). It also appears that AK3L1 is a
predicted target of miR-195, according to Targetscan and miRanda, two miR target
prediction softwares with miR-195 being a micro-RNA that attracted my interest in
Chapter 4.
The results of the complementation assay with a mix of 2 silencing
constructs for AK3L1, namely human GIPZ lentiviral shMiR V2LHS_59300 and human
GIPZ lentiviral shMiR V2LHS_59298 show a rescue compared to the negative control
(Figure 4.5) and at a slightly higher level than the one with silencing CLCA2. A repeat
experiment showed a rescue at an even higher level, with approximately 30% more
growing colonies.
4.2.2.3 TRIB2 silencing bypassed senescence
Tribbles homolog 2 (Trib2) was up-regulated upon senescence in this study but was
previously identified as a down-regulated transcript in leukemic cells undergoing nonsenescence growth arrest. In mechanistic studies, Trib2 was identified as an oncogene
with pro-proliferation properties in prostate cancer progression and acute myeloid
leukemia, the latter effect being mediated through regulation of the C/EBP family of
proteins and notably inactivation of cEBPalpha and cEBPbeta (Keeshan, He et al. 2006;
Naiki, Saijou et al. 2007). This is not in agreement with my finding that TRIB2 was upregulated upon senescence and that this up-regulation was accompanying the growth
arrest suggesting anti-growth properties rather than proliferative properties in the
CL3EcoR cells.
169
Figure 4.5: In vitro validation of up-regulated microarray targets by silencing constructs
(A) Silencing of AK3L1 and CLCA2: CL3EcoR cells were infected in triplicate with a mix of lentiviruses
shRNAmir silencing constructs expressing AK3L1 (V2LHS_59300 and V2LHS_59298) and CLCA2
(V2LHS_197853, V2LHS_197750 and V2LHS_199854) and assayed for growth complementation at
38°C. After 3 weeks the number of growing colonies were counted. (B) Silencing of TRIB2, CDKN2A,
DAPK1, BLCAP and RUNX1: CL3EcoR cells were infected in triplicate with a mix of lentiviruses
shRNAmir silencing constructs expressing TRIB2 (V2LHS_200999 and V2LHS_200588), CDKN2A
(V2LHS_195839, V2LHS_200698 and V2LHS_200168), DAPK1 (V2LHS_62089, V2LHS_62085 and
V2LHS_62084),
BLCAP(V2LHS_90065
and
V2LHS_90063)
and
RUNX1(V2LHS_150257,
V2LHS_150259 and V2LHS_150256) and assayed for growth complementation at 38°C. After 3 weeks
the number of growing colonies were counted.
170
In addition, I found that although TRIB2 was up-regulated, not only cEBPalpha and beta
expression were not down-regulated but cEBPbeta was actually up-regulated upon
senescence.
The results of the complementation assay with a mix of 2 lentiviral silencing constructs
for TRIB2 namely human GIPZ lentiviral shMiR V2LHS_200999 and human GIPZ
lentiviral shMiR V2LHS_200588 permitted the rescue from senescence of the cells
when compared to the negative control (Figure 4.5).
4.2.2.4 CDKN2A silencing bypassed senescence
CDKN2A or p16INK4A has been linked tightly to the senescence pathways (see
introduction chapter). This gene is known to be an important tumour suppressor gene
capable of inducing cell cycle arrest. The transcript of p16INK4A contains an alternate
open reading frame (ARF) that functions as a stabilizer of the tumour suppressor protein
p53 as it can bind with MDM2 and blocks its nucleo-cytoplasmic shuttling by
sequestering it in the nucleolus which end up blocking MDM2-induced degradation of
p53 thereby enhancing p53-dependent transactivation and apoptosis. ARF can also
trigger G2 growth arrest and apoptosis in a p53-independent manner by preventing the
activation of cyclin B1/CDC2 complexes.
CDKN2A was found here to be up-regulated upon senescence in the HMF3A. However,
it is important to note that we previously failed in silencing CDKN2A with various
different silencing constructs (see introduction chapter) and the expression has not been
checked on this occasion. Nevertheless, the potential silencing of p16INK4A with a mix of
3 lentiviral silencing constructs namely human GIPZ lentiviral shMiR V2LHS_195839,
human GIPZ lentiviral shMiR V2LHS_200698 and human GIPZ lentiviral shMiR
V2LHS_200168 seem to have a rescuing effect here at a level equivalent to TRIB2
(Figure 4.5).
171
4.2.2.5 DAPK1 silencing was not sufficient to bypass
senescence
Death-associated protein kinase 1 is a positive mediator of gamma-interferon induced
programmed cell death (Deiss, Feinstein et al. 1995; Shohat, Shani et al. 2002; Bialik
and Kimchi 2004). It is a unique multidomain kinase acting both as a tumour suppressor
and an apoptosis inducer. TCR-induced NF-κB activation was also shown to be
activated as a target of DAPK (Chuang, Fang et al. 2008).
It was also up-regulated in my data, and therefore was tested in a complementation assay
using a mix of 3 lentiviral silencing constructs of DAPK1.
Here, the complementation assay was not a success and DAPK1 silencing with a mix of
3 lentiviral silencing constructs for DAPK1 namely human GIPZ lentiviral shMiR
V2LHS_62089, human GIPZ lentiviral shMiR V2LHS_62085 and human GIPZ
lentiviral shMiR V2LHS_62084 was not able to bypass the conditional cell cycle arrest
(Figure 4.5).
4.2.2.6 BLCAP silencing bypassed senescence at a low
level
BLCAP was identified as a tumour suppressor protein that reduces cell growth by
stimulating apoptosis. HeLa cells expressing BLCAP show reduced cell growth
compared to vector-transfected cognate cells and this expression also led to growth
arrest and significantly enhanced apoptosis in vitro and reduced tumour formation in
vivo (Zuo, Zhao et al. 2006). Over-expressed BLCAP resulted in growth inhibition of a
human tongue cancer cell line Tca8113 in vitro, accompanied by S phase cell cycle
arrest and apoptosis. Taken together, BLCAP may play a role not only in regulating cell
proliferation but also in coordinating apoptosis and cell cycle via a novel way
independent of p53 and NF-κB (Yao, Duan et al. 2007).
172
BLCAP is up-regulated upon senescence-like growth arrest and it was therefore not
surprising that its up-regulation was linked to growth inhibition in the literature. For
these reasons, BLCAP silencing was investigated in a complementation assay.
A mix of 2 lentiviral silencing constructs of BLCAP, namely human GIPZ lentiviral
shMiR V2LHS_90065 and human GIPZ lentiviral shMiR V2LHS_90063 was used for
complementation assay and showed a rescue of the cells from senescence when
compared to the negative control (Figure 4.5).
4.2.2.7 RUNX1 bypassed senescence
All three family members: RUNX1, 2 and 3 possess the ability to induce senescence-like
growth arrest in primary murine fibroblasts (Linggi, Muller-Tidow et al. 2002; Wotton,
Blyth et al. 2004; Kilbey, Blyth et al. 2007). An analogous role was suggested for Runx1
in hematopoietic progenitors by the failure of NRAS- induced growth suppression in
cells lacking Runx1 (Motoda, Osato et al. 2007).
For these reasons and because RUNX1 was also up-regulated upon senescence, RUNX1
silencing was tested with a mix of 3 lentiviral silencing constructs namely human GIPZ
lentiviral shMiR V2LHS_150257, human GIPZ lentiviral shMiR V2LHS_150259 and
human GIPZ lentiviral shMiR V2LHS_150256 in a complementation assay in the
CL3ECoR cells. The results show a rescue compared to the negative control (Figure 4.5).
4.2.2.8 GRAMD3 silencing bypassed senescence
GRAMD3 is one of the genes up-regulated upon senescence. It also appears that
GRAMD3, according to Targetscan and miRanda, two miR target prediction softwares,
is a potential target of miR-195 and miR-25, two of the micro-RNA studied in chapter 4.
No literature was available for GRAMD3. However, its silencing by 2 individual
lentiviral
silencing
constructs
namely
Human
GIPZ
lentiviral
shRNAmiR
V2LHS_235566 and Human GIPZ lentiviral shRNAmiR V2LHS_135659 showed to
173
bypass the growth arrest in a reproducible manner and at a level with a number of
colonies stained after 3 weeks above 100 (Figure 4.6). This experiment was repeated
with the same results.
4.2.2.9 SCN2A silencing was not sufficient to bypass
senescence
SCN2A stands for sodium channel, voltage-gated, type II, alpha subunit. Voltage-gated
sodium channels (NaV) are responsible for action potential initiation and propagation in
excitable cells, including nerve, muscle, and neuroendocrine cell types. They are also
expressed at low levels in non-excitable cells, where their physiological role is unclear.
SCN2A was one of the 20 top up-regulated genes upon senescence.
Its silencing was tested with 2 individual silencing constructs namely human GIPZ
lentiviral shMiR V2LHS_202838 and human GIPZ lentiviral shMiR V2LHS_203129.
The results show a weak rescue but with a number of colonies higher above the Lamin
background. The experiment was repeated with similar results (Figure 4.6). This result
was not considered conclusive enough.
4.2.3 Down-regulated genes upon senescence: Does ectopic expression bypass
the growth arrest?
James Robinson, a BSc rotation student, tried to obtain antibodies to verify the
expression for all these proteins, namely HMGB2, DEPDC1, NEK2 and MLF1-IP 88
and 401, and to carry out Western blotting but none of the available antibodies except
for MELK and FOXM1 worked.
174
Figure 4.6:Silencing of GRAMD3 and SCN2A
CL3EcoR cells were infected in triplicate with lentiviruses expressing the indicated shRNAmir silencing
constructs and assayed for growth complementation at 38°C. After 3 weeks the number of growing
colonies were counted.
175
4.2.3.1 HMGB2
HMGB2 encodes for a member of the non-histone chromosomal high mobility group
protein family. In vitro studies have demonstrated that this protein is able to efficiently
bend DNA and form DNA circles. Previous reports have shown that architectural DNAbending/looping chromosomal proteins HMGB1 and HMGB2 (formerly known as
HMG1 and HMG2), which function in a number of biological processes including
transcription and DNA repair, interact in vitro with p53 and stimulate p53 binding to
DNA containing p53 consensus sites (Jayaraman, Moorthy et al. 1998; Brickman, Adam
et al. 1999; Imamura, Izumi et al. 2001). HMGB1 and 2 were also shown to physically
interacts with two splicing variants of p73, alpha and beta and stimulate p73 binding to
different p53-responsive elements and therefore modulate its activity (Stros, Ozaki et al.
2002).
HMGB2 was also strongly down regulated upon senescence; therefore, it was chosen to
be tested by complementation assay.
PLPC-HMGB2, a full length expression construct for HMGB2, was packaged with
ecotropic phoenix cells to produce retroviral supernatant which was used for
complementation assay in the CL3EcoR cells. After selection, cells were reseeded as usual
and placed at 38°C for 3 weeks. The cells did not rescue above the background (plpcx
empty vector) (data not shown). A repeat of this experiment showed the same result.
However, because the expression of HMGB2 was not verified, it is not possible to
conclude on the actual efficiency of its ectopic expression.
4.2.3.2 DEPDC1
DEPDC1 was shown to be up-regulated in bladder cancer cells. In addition, suppression
of DEPDC1 expression with small-interfering RNA significantly inhibited growth of
bladder cancer cells (Kanehira, Harada et al. 2007). It was also represented 3 times (3
176
different oligos corresponding to the same gene) in the top 25 down-regulated targets
upon senescence.
For these reason, ectopic expression of this gene was also tested by complementation
assay with retroviral expression of PLPC-DEPDC1, a full length expression construct
for DEPDC1.
The results showed rescue compared to the negative control (Figure 4.7). However,
these results were not reproducible at a satisfactorily level of rescue with very weak
rescue for some of the cultures tested. The effect of DEPDC1 on senescence growth
arrest remains, therefore, to be confirmed. In addition, the expression of DEPDC1 was
not verified and therefore, the actual efficiency of its ectopic cannot be assessed.
4.2.3.3 BUB1B
Bub1 is a kinase believed to function primarily in the mitotic spindle checkpoint.
Mutation or aberrant Bub1 expression is associated with chromosomal instability,
aneuploidy, and human cancer (Cahill, Lengauer et al. 1998). Bub1 expression was
reported to be low in cells undergoing replicative senescence. It was also described that
targeting Bub1 by RNAi or simian virus 40 (SV40) large T antigen in normal human
diploid fibroblasts results in premature senescence.
Premature senescence caused by lower Bub1 levels was dependant on p53 as senescence
induction was blocked by dominant negative p53 expression or depletion of
p21CIP1/WAF1/Sdi1, a p53 target (Gjoerup, Wu et al. 2007; Gao, Ponte et al. 2009).
Since BUB1 was in the top 25 down regulated targets upon senescence in the CL3 EcoR
cells with two different oligos, highlight the quality of the microarray and the
importance of BUB1 in the senescence processes. All together, this makes BUB1 a
perfect target to ectopically express in the conditional system.
177
Figure 4.7: In vitro validation of microarray down-regulated targets by ectopic expression
(A) Ectopic expression of MLF-IP88, MLF-IP401, DEPDC1 and NEK2: CL3EcoR cells were infected
in triplicate with retroviruses expressing the indicated PLPC expression constructs and assayed for growth
complementation at 38°C. After 3 weeks the number of growing colonies were counted. (B) Ectopic
expression of hBUB1: CL3EcoR cells were infected in triplicate with retroviruses expressing the indicated
PLPC expression constructs and assayed for growth complementation at 38°C. After 3 weeks the number
of growing colonies were counted
178
pLB(N)C- HA-BUB1 was packaged into amphotropic phoenix cells and used to infect
the HMF3A cells. After blastocidin selection for 15 days, the cells were reseeded at
0.5x105 in T-75 cm2 and shifted to 38C for 3 weeks before being stained as described
previously. The results show a clear rescue of the cells, although at a low level,
compared to the negative control.
In addition, because HA-BUB1B expression had previously been shown by Ole Gjoerup
to be functional, I was confident that the expression construct of BUB1B worked and
therefore that BUB1B was causal to senescence escape.
4.2.3.4 NEK2
Nek2 is a cell-cycle-regulated protein kinase that localizes to the centrosome and
is likely to be involved in regulating centrosome structure at the G(2)/M transition. Nek2
is expressed as two splice variants. These isoforms, designated Nek2A and Nek2B, are
detected in primary blood lymphocytes as well as adult transformed cells (Hames and
Fry 2002). Expression levels of the Nek2 kinase in human cancer cell lines and primary
tumours revealed that Nek2 protein is elevated 2- to 5-fold in cell lines derived from a
range of human tumours including those of cervical, ovarian, breast, prostate, and
leukemic origin (Hayward, Clarke et al. 2004). More recently, NEK2 was also reported
by the same group to be abnormally expressed in a wide variety of human cancers
(Hayward and Fry 2006).
NEK2 was also down-regulated in the HMF3A conditional system and its ectopic
expression was chosen to be tested by complementation assay using a full length
expression construct.
The results show that NEK2 expression was unable to rescue the cells from senescence
(Figure 4.7). However, again its ectopic expression was not confirmed so no conclusion
can be made on its actual efficiency to bypass senescence.
179
4.2.3.5 MELK
Several studies found a correlation between MELK expression and the malignancy of
several cancers. MELK has been shown to be over-expressed by at least a 5-fold
increase in invasive glioblastoma multiforme (GBM). In addition, in the examination of
more than 100 tumours of the central nervous system, progressively higher expression of
MELK was found to correlate with astrocytoma grade. Similar level of over-expression
was also observed in medulloblastoma. Furthermore, MELK knockdown in malignant
astrocytoma cell lines caused a reduction in proliferation and anchorage-independent
growth in in vitro assays (Marie, Okamoto et al. 2008). Melk was also found highly
expressed in murine neural stem cells and regulated their liferation and correlated with
pathologic grade of brain tumours. In primary cultures from human glioblastoma and
medulloblastoma, MELK knockdown by siRNA results resulted in inhibition of the
proliferation and survival of these tumours (Nakano, Masterman-Smith et al. 2008).
Using accurate genome-wide expression profiles of breast cancers, another study found
MELK to be significantly over-expressed in the great majority of breast cancer cells.
Suppression of MELK expression by small interfering RNA significantly inhibited
growth of human breast cancer cells (Lin, Park et al. 2007). Altogether, these results
suggested a critical role for MELK in cell proliferation and tumourigenesis.
For these reasons, MELK ectopic expression was tested by complementation assay and
the results (data not shown) showed that introduction of MELK in the HMF3A cells was
not sufficient to bypass the conditional senescence. However, western blotting analysis
was performed by James Robinson with a MELK antibody (data not shown) revealed
that MELK was not expressed at a sufficient level in these cells and therefore since the
expression of MELK in these cells was not confirmed, I could not conclude on the effect
of MELK expression itself on the growth arrest.
The ectopic expression of MELK would be definitely worth investigating further
perhaps with a different expression construct but, in reason of the short time scale was
dropped.
180
4.2.3.6 MLF1-IP two splice forms 88 and 401
The myelodysplasia/myeloid leukemia factor 1-interacting protein MLF1-IP is a novel
gene which encodes for a putative transcriptional repressor. MLF1-IP has been shown to
be over-expressed in human and rat glioblastoma (GBM) especially in the tumour core
where it was co-localized with MLF1 and nestin (Hanissian, Teng et al. 2005).
In biological studies, there have been several observations suggesting that MLF1 is
physiologically involved in a tumour suppressor pathway. MLF1 has been found to be
over-expressed in more than 25% of myelodysplasic syndromes (MDS) -associated
cases of AML, in the malignant transformation phase of MDS, and in lung squamous
cell carcinoma (Matsumoto, Yoneda-Kato et al. 2000; Sun, Zhang et al. 2004). The
aberrant over-expression is usually related to mutations and to inactivation of p53 in
various cell lines (Yoneda-Kato, Tomoda et al. 2005). It was also reported that MLF1 is
a negative regulator of cell cycle progression that functions upstream of the tumour
suppressor p53 (Dornan, Wertz et al. 2004). The introduction of NPM-MLF1 into earlypassage murine embryonic fibroblasts allowed the cells to escape from cellular
senescence at a markedly earlier stage and induced neoplastic transformation in
collaboration with the oncogenic form of Ras (Yoneda-Kato and Kato 2008). MLF1-IP
also happened to be down-regulated upon senescence.
For this reason, MLF1-IP was a good target to try and express ectopically in a
complementation assay. The cDNA was cloned in LPCX after amplification by Pr.
Parmjit Jat and these were used for the complementation as described previously.
Colonies numbers and phenotype were not convincing enough to indicate a significant
bypass of senescence with either isoform expressions (Figure 4.7A). In addition, the
expression of MLF1-IP was not verified making its actual efficiency impossible to
assess.
181
4.2.3.7 DBF4, CDKN2C (p18) and PLK4
DBF4, CDKN2C (p18) and PLK4 were the other targets chosen due to their
involvement in cell proliferation and their down-regulation upon senescence. Full length
ectopic expression constructs were generated but, these did not yield sufficient number
of puromycin resistant clones. So after a few tries and due to the timescale, their
investigation was dropped.
4.2.3.8 FOXM1
FOXM1 is a transcription factor that belongs to the evolutionarily conserved Forkhead
family comprising more than 50 transcription factors that share a conserved Forkhead or
Winged–helix DNA-binding domain (Laoukili, Stahl et al. 2007; Myatt and Lam 2007;
Wierstra and Alves 2007). In humans, there are 17 Fox gene subfamilies (FOXA-R)
with at least 41 different genes. Despite the highly conserved Forkhead DNA binding
domain (DBD), the function and regulation of the FOX proteins varies considerably
between the different families probably due to sequence variations outside the DBD
allowing for functional diversity and regulation. FOX protein family members play a
role in a wide variety of biological processes such as development, differentiation,
proliferation, apoptosis, migration, invasion and ageing; some such as FOXM1 have
even been linked to cancer (Myatt and Lam 2007).
Human FOXM1 exists as three splice variants: FOXM1 a, b and c (A, B, C in Figure
4.8A). All three isoforms bind to the same DNA sequences but only FOXM1b and
FOXM1c are transactivators. Disruption of the transactivation domain but retention of
the DBD in FOXM1a indicates that it has the potential to be a naturally occurring
dominant negative variant (Figure 4.8A).
182
A
183
B
Figure 4.8: FOXM1
(A) FOXM1 splice variants: DNA gene containing 10 exons, 2 of which being splicing exons Va (A1)
and VIIa (A2) that originates 3 different splice variants, encoding for 3 FOXM1 protein isoforms:
FOXM1a, containing both alternative exons, FOXM1b, not containing any alternative exons and
FOXM1c only containing exon Va. (B) Involvement of the FoxM1-regulatory gene network in the
regulation of cell cycle progression and maintenance of the genomic stability (from Laoukili, 2007).
Several microarray analyses studies have revealed numerous FoxM1 target genes. The most significant of
these genes can be clustered in function of their role in the regulation of the cell cycle, more specifically of
the G2/M-phases of the cell cycle: mitotic entry, mitotic spindle checkpoint and/or chromosome
segregation, and cytokinesis and mitotic exit. The proper coordination of the expression of these genes in
space and time participates to proper cell cycle progression and maintenance of the genomic stability.
184
FOXM1 exhibits a proliferation-specific expression pattern (Wierstra and Alves 2007).
It is highly expressed in the developing embryo but is turned off upon terminal
differentiation. In the adult expression is limited to proliferating cell types and selfrenewing tissues such as thymus, testis, small intestine and colon that contain
proliferative cells; significantly lower levels are found in ovary, spleen and lung.
FOXM1 expression also decreases upon ageing. FOXM1 expression has been detected
in all proliferating cells but is not expressed in quiescent or terminally differentiated
cells. However FOXM1 is readily induced when quiescent cells re-enter the cell cycle
upon stimulation. The increase in FOXM1 expression is initiated in late G1 at the onset
of S phase reaching a maximal level which is maintained throughout G2 and mitosis.
However the transcriptional activity of FOXM1 is only maximal during G2 and
correlates with its increased phosphorylation. During exit from mitosis, FOXM1 is
actively degraded by the anaphase-promoting complex (Laoukili, Alvarez-Fernandez et
al. 2008; Park, Wang et al. 2008).
FOXM1 contains an N-terminal auto-repressor domain that inhibits transactivation by an
intramolecular interaction with the C-terminal transactivation domain (TAD) (Wierstra
and Alves 2007; Park, Wang et al. 2008).
This repression can be relieved by
phosphorylation of multiple cdk sites within the TAD; cyclinA/cdk2 has been suggested
to be essential for phosphorylation of these sites (Laoukili, Alvarez-Fernandez et al.
2008). Phosphorylation by cyclinE/cdk2 and PLK1 has also been suggested to be
important for regulating FOXM1 activity. CyclinD/cdk4, 6 may activate FOXM1
indirectly by relieving it‘s inhibition by the retinoblastoma protein. Phosphorylation by
MAPK has been proposed to be required for translocation of FOXM1 to the nucleus
(Ma, Tong et al. 2005). Expression studies have indicated that FOXM1 regulates
expression of the G2-specific gene expression signature of mammalian cells (Laoukili,
Kooistra et al. 2005; Wang, Chen et al. 2005; Mooi and Peeper 2006). This comprises
Cyclin B, Polo-like-kinase 1 (PLK1), Aurora B, Cdc25B, CENP-F, NEK2 and many
other regulators of cell cycle progression and genomic stability (Figure 4.8B).
185
Since many of these genes as well as FOXM1 are down-regulated in our HMF3A cells
when they undergo senescence, we initiated a collaboration with Rene Medema to
determine if FOXM1 has a causative role in this process. They provided us with
lentiviral expression constructs for full length FOXM1c (FOX WT), a sumoylationdefective inactive mutant (FOX 6K, not published) and a constitutively active, nondegradable, N-terminal deleted FOXM1c (FOXN, comprising amino acids 210763; Laoukili et al, 2008).
To confirm the microarray, the levels of FOXM1 were analysed by Western blot by
James Robinson with an anti-FOXM1 antibody in either growing cells after 7 days at
34°C or in senescent cells shifted to 38°C for one week. The western blot membrane
was scanned and the intensity was calculated for each of the bands. The absolute
intensity was then calculated by the ratio FOXM1/B2-microglobulin. The results show a
decrease in the level of FOXM1 protein in senescent cells in all three experiments
although the exact level of expression of FOXM1 varied (Figure 4.9). These results
corroborate the microarray analysis and confirm a decrease of FOXM1 levels in
senescent cells.
The FOXM1 inserts within the three lentiviral constructs were excised and recloned into
PLPCX by Catia Caetano and these were used in a complementation assay. The staining
result (Figure 4.10) showed that constitutively active FOXM1 (FOXM1NKEN)
abrogates senescence in our HMF3A cells upon temperature shift whereas full length
FOXM1 (FOXM1) or the mutant (FOXM1-6K) were unable to rescue even when highly
expressed. The experiment was repeated with similar results although with a lower
number of growing colonies.
186
Figure 4.9: FOXM1 protein expression
The levels of FOXM1 were analysed by Western blot with an anti-FOXM1 antibody in either growing
cells after 7days at 34°C or in senescent cells shifted to 38°C for 3 weeks. The western blot membrane
was scanned and the intensity was calculated for each of the bands. The absolute intensity was then
calculated by the ratio FOXM1/B2-microglobulin. This Figure represents 3 Western Blot repeats.
187
Figure 4.10: Ectopic expression of FOXM1 WT, FOXM1ΔNΔKEN and FOXM1-6K :
EcoR
CL3
cells were infected in triplicate with retroviruses expressing the indicated PLPC expression
constructs and assayed for growth complementation at 38°C. After 3 weeks the number of growing
colonies were counted.
188
In addition, duplicate cultures expressing the three different FOXM1 constructs
as well as a negative control, PLPCX, were grown at 34°C and analysed by western blot
with an anti-FOXM1 antibody The results showed an expression of FOXM1 in the three
conditions (Figure 4.11, lane 1, 2 and 3) but not in the negative control (Figure 4.11,
lane 4). It is possible to note that the constitutively active mutant of FOXM1,
FOXNKEN, show a lower band at 75KDa, corresponding to the size of the truncated
protein (Figure 4.11, lane 2).
4.3 NF-κB PATHWAY ACTIVATION UPON SENESCENCE IS CAUSAL TO
SENESCENCE
4.3.1 Objectives
In silico promoter analysis of differentially expressed genes upon senescence suggested
that the NF-κB pathway may be activated upon irreversible growth arrest (Hardy,
Mansfield et al. 2005). Additionally, simple observation of the microarray data
suggested similar conclusions. The objectives in this section were to investigate further
the involvement of NF-κB pathway in senescence and use different approaches to
validate in vitro my initial hypothesis.
4.3.2 NF-κB pathway is activated upon senescence at the mRNA level
Microarray analysis can provide very useful information on pathways and pattern. To
determine whether the NF-κB pathway was actually involved in senescence, two
methods were employed: The first one was to analyse the 4 transcription factor motifs of
NF-κB in all the differential gene data set, the second was to extract all the genes that
are known targets of NF-κB from the differential gene data set and to look at their
modulation.
189
Figure 4.11: FOXM1 protein expression in cells expressing FOXM1 WT, FOXM1ΔNΔKEN and
FOXM1-6K
The levels of FOXM1 were analyzed by Western blot with an anti-FOXM1 antibody in the cells
expressing the indicated PLPC expression constructs.
190
4.3.2.1 Transcription factor motif matrix module
To determine if the differentially expressed genes contained binding sites for the family
of NF-κB transcription factors, this expression dataset was compared with a motif
module map comprising a matrix of 12,254 genes and 2,394 known transcription factors
(http://motifmap.googlepages.com) (Adler, Lin et al. 2006). All four NF-κB factor
motifs were found to be present within the promoters of the up- and down-regulated
genes (Table 4.5). In the up-regulated genes the NF-κB motifs were present within the
promoters of 200, 134, 124 and 114 genes, ranking them in the top 3%, 5.7%, 6.4% and
7.4% abundant motifs respectively. In the down-regulated genes, the NF-κB motifs were
present within the promoters of 217, 175, 144 and 116 genes, ranking them in the top
4.2%, 5.9%, 7.5% and 10.3% abundant motifs respectively.
4.3.2.2 NF-κB targets gene expression modulation
To determine if the differential GA gene set comprised known NF-B targets, the list of
differentially expressed genes was compared to the set of 960 putative NF-B targets
proposed by Gilmore (http://www.nf-kb.org).
93 NF-B targets were found to be
differentially expressed; 67 of these were up-regulated (Table 4.6A and Supplementary
Table S4.7 (supplementary on a CD)) whereas 26 were down-regulated (Table 6B and
Supplementary Table S4.7). IL1A and IL1B were the most highly up-regulated genes.
The senescence-associated secretory phenotype (SASP) cytokine IL6 was also upregulated. The other SASP cytokines IL8 and IGFBP7 were also up-regulated (Table
4.6C) but since the adjusted P-values were greater than 0.00001, they were not identified
as significant. Interestingly IL6 expression was up-regulated to a greater extent by
serum starvation. Almost all of the top NF-B targets found to be up-regulated upon
senescence growth arrest were also up-regulated upon serum starvation including IL1A
and B, the most highly up-regulated NF-B targets; generally the modulation by serum
starvation was greater than with senescence growth arrest.
191
UP
V_NFKB_Q6_01.wtmx
No_of_genes Position out of 2394
TF
200
73
Rank by percent
3.0
V_NFKAPPAB65_01.wtmx
134
136
5.7
V_NFKAPPAB_01.wtmx
124
154
6.4
V_NFKAPPAB50_01.wtmx
114
178
7.4
DOWN
V_NFKB_Q6_01.wtmx
No_of_genes Position out of 2394
TF
217
100
Rank by percent
4.2
V_NFKAPPAB65_01.wtmx
175
142
5.9
V_NFKAPPAB_01.wtmx
144
180
7.5
V_NFKAPPAB50_01.wtmx
116
246
10.3
Table 4.5: Transcription factor motifs
The differential expression dataset was compared with a motif module map comprising a matrix of
12,254 genes and 2,394 known transcription factors (Adler et al. 2006). All four NF-κB factor motifs were
found to be present within the promoters of the up- and down-regulated genes. In the up-regulated genes
the NF-κB motifs were present within the promoters of 200, 134, 124 and 114 genes, ranking them in the
top 3%, 5.7%, 6.4% and 7.4% abundant motifs respectively. In the down-regulated genes, the NF-κB
motifs were present within the promoters of 217, 175, 144 and 116 genes, ranking them in the top 4.2%,
5.9%, 7.5% and 10.3% abundant motifs respectively.
192
A
B
Symbol logFC GA logFC Q logFC HS
logFC
wt_LT
logFC
GSE_p53
logFC
logFC E1A logFC E7 E2F-DB
logFC
pRS_p53
logFC
pRS_p21
IL1A
3.42
3.82
0.22
-4.36
-3.28
-5.18
-1.05
-0.04
-0.70
-0.26
IL1B
3.33
3.54
0.06
-4.60
-4.13
-5.44
-1.98
-0.86
-1.41
-1.48
IL1B
3.29
3.43
0.38
-4.58
-3.83
-5.89
-1.82
-0.80
-1.31
-1.37
BMP2
2.64
4.73
0.24
-5.09
-3.59
-4.51
-3.45
-2.96
-1.45
-1.91
BMP2
2.25
4.80
0.24
-5.00
-3.18
-4.42
-3.51
-3.20
-1.47
-1.80
SOD2
2.10
3.04
0.72
-2.27
-2.00
-2.83
-0.92
-0.63
-0.71
-0.87
40118
2.04
2.14
-0.56
-2.97
-2.23
-2.96
-1.84
-1.41
-1.74
-1.71
IL6
2.01
4.16
0.13
-3.37
-2.65
-5.35
-2.64
-1.97
-1.24
-1.93
AKR1C1
2.01
3.30
-0.65
-2.44
-0.82
-4.22
-1.2
-1.19
0.46
-0.12
TNFAIP3
1.96
2.42
0.32
-3.27
-3.30
-4.35
-1.64
-1.60
-1.68
-1.74
IL32
1.96
0.78
0.17
-1.42
-1.23
-0.26
-0.99
-0.36
-0.86
-1.11
40118
1.93
2.10
-0.38
-2.89
-2.36
-2.67
-1.79
-1.34
-1.67
-1.76
40118
1.92
1.96
-0.61
-2.74
-2.13
-2.52
-1.53
-1.11
-1.55
-1.51
CCL2
1.92
3.55
0.62
-1.14
-1.75
-5.25
-1.06
-1.65
-1.05
-1.26
TNFAIP3
1.90
2.53
0.31
-3.58
-3.42
-4.90
-1.74
-1.77
-1.68
-1.90
CCL20
1.84
3.63
0.16
-2.96
-3.49
-2.34
-2.02
-1.37
-2.07
-2.61
CSF2
1.81
2.97
-0.29
-3.41
-3.31
-3.97
-0.77
0.94
-1.07
-1.23
AKR1C1
1.79
4.57
-0.28
-3.22
-1.47
-4.07
-2.14
-2.19
0.12
-0.94
AKR1C1
1.78
3.16
-0.38
-2.03
-0.73
-2.80
-1.11
-1.12
0.50
-0.08
FTH1
1.75
1.30
-0.01
-1.28
-0.62
-2.96
-0.64
-0.65
-0.65
-0.44
SOD2
1.69
3.53
0.85
-2.12
-2.49
-2.84
-0.91
-0.72
-0.54
-0.83
CDKN1A
1.68
0.73
-0.66
-1.54
-2.65
-0.40
-0.39
-0.35
-1.95
-1.17
GCLC
1.66
1.86
-0.68
-1.58
-1.42
-1.36
-1.35
-1.5
-1.55
-1.71
ANGPT1
1.64
-1.04
0.06
-0.98
0.48
-5.61
-1.03
-1.88
-0.02
-0.13
logFC
wt_LT
logFC
GSE_p53
logFC
pRS_p53
logFC
pRS_p21
Symbol logFC GA logFC Q logFC HS
logFC E1A logFC E7 logFC DB
BRCA2
-2.20
0.32
0.05
1.88
1.08
2.04
1.56
1.98
1.9
1.65
DPYD
-1.50
-2.15
-0.16
0.25
0.39
0.04
0.58
0.19
0.53
0.27
BRCA2
-1.42
0.63
-0.10
1.15
1.09
1.19
0.99
1.37
0.88
0.73
UCP2
-1.33
-0.67
-0.02
1.38
0.65
1.82
0.73
1.39
-0.05
0.04
S100A10
-1.18
-0.25
-0.37
0.99
0.23
1.73
0.63
0.90
0.38
0.47
TWIST1
-1.03
0.46
0.28
1.51
1.45
0.48
0.65
0.41
0.8
0.46
CD44
-0.99
0.49
-0.73
0.66
0.65
-1.08
0.44
0.44
0.65
0.78
PPP5C
-0.98
-0.04
-0.33
0.56
0.42
0.54
0.13
0.18
0.45
0.33
HOXA9
-0.98
-0.26
-0.40
0.74
1.29
1.78
0.16
0.08
0.49
0.78
HOXA9
-0.91
-0.16
-0.35
0.77
1.36
1.85
0.11
0.15
0.46
0.68
TNC
-0.79
0.36
-0.09
-0.20
0.10
-3.97
-0.51
-1.37
-0.22
-0.25
EGFR
-0.79
-0.40
-0.33
-0.12
0.34
-0.83
-0.27
-0.12
0.25
-0.08
PIM1
-0.76
0.56
0.34
1.18
0.99
0.38
0.36
-0.05
0.46
0.27
AHCTF1
-0.74
0.35
0.04
0.39
0.90
0.26
0.12
0.43
0.47
0.32
AHCTF1
-0.73
-0.02
-0.07
0.61
0.53
0.67
0.43
0.59
0.49
0.41
EGFR
-0.72
-0.06
-0.34
-0.02
0.59
-1.09
-0.41
-0.29
0.10
-0.28
BMI1
-0.70
-0.28
-0.18
0.18
0.43
0.89
0.27
0.31
0.17
0.18
NR3C1
-0.69
-0.14
-0.02
-0.07
0.32
-0.19
-0.29
-0.49
0
0.02
UBE2M
-0.69
0.01
-0.22
0.18
0.15
0.10
-0.09
0.09
0.03
0.17
GNB2L1
-0.65
0.25
0
0.61
0.50
1.09
0.44
0.44
0.35
0.30
PTEN
-0.60
-0.32
-0.25
0.57
0.46
0.48
0.49
0.40
0.56
0.58
HMGN1
-0.57
-0.12
0.03
0.63
0.63
0.82
0.58
0.64
0.43
0.36
DPYD
-0.55
-1.22
-0.29
0.12
-0.01
0.04
0.53
0.22
0.46
0.25
AHCTF1
-0.55
-0.10
-0.10
0.58
-0.01
0.77
0.61
0.72
0.65
0.70
Table 4.6: Senescence specific changes in NF-κB target genes expression with complementation Log2
fold changes in gene expression that occur upon growth arrest (GA), heat shock (HS) and upon serum
starvation (Q). If changes in gene expression are specific for the senescence growth arrest, they should be
reversed upon its abrogation. Up-regulated transcripts are indicated in green whereas down-regulated
transcripts are in red. Results for the top 24 up- (A) and down-regulated (B)
193
C
logFC GA logFC Q logFC HS
logFC wt logFC
logFC
logFC
LT
pRS p53 GSE-p53 pRS p21
logFC
E1A
logFC
logFC E7
E2F-DB
ID
Symbol
201783_s_at
RELA
-0.37
0.08
-0.33
-0.09
-0.27
0.06
-0.52
-0.31
-0.44
-0.34
230202_at
RELA
0.1
-0.16
-0.21
-0.07
-0.19
-0.02
-0.1
-0.4
-0.14
-0.13
209878_s_at
RELA
-0.43
-0.03
-0.34
0.1
0.1
-0.1
-0.1
-0.08
-0.27
-0.1
205205_at
RELB
0.29
1.45
0.33
-0.77
-0.56
-1.01
-0.58
-0.81
-0.47
-0.6
209239_at
NFKB1
-0.35
1.17
0.23
-0.01
-0.03
-0.14
-0.26
0.24
0.34
0.14
207535_s_at
NFKB2
0.45
1.15
0.19
-0.4
0.08
-0.55
-0.08
-0.28
-0.01
-0.09
209636_at
NFKB2
0.39
0.9
0.14
-0.48
-0.05
-0.73
-0.21
-0.33
0.06
0.02
211524_at
NFKB2
-0.08
-0.1
-0.15
0.12
0.09
-0.11
0.17
0.18
0.06
0.18
212312_at
BCL2L1
0.65
-0.24
-0.49
-1.24
-0.89
-1.2
-0.02
-1.19
-0.12
-0.49
206665_s_at
BCL2L1
0.59
-0.24
-0.62
-0.81
-0.68
-0.6
0.01
-1.21
0.02
-0.4
215037_s_at
BCL2L1
0.6
-0.54
-0.47
-0.91
-0.54
-1.05
0.17
-1.03
0.14
-0.12
231228_at
BCL2L1
-0.04
0.02
-0.13
-0.14
-0.11
-0.4
0.04
-0.18
0.22
0.14
201236_s_at
BTG2
1.83
0.22
-0.2
-1.37
-2.2
-1.92
0.46
0.01
0.01
-0.09
201235_s_at
BTG2
0.73
-0.05
-0.05
-0.41
-0.64
-0.26
0.37
0.03
0.12
-0.07
223710_at
CCL26
1.81
-0.1
0.19
-0.43
0.54
0.59
-0.37
-2.4
-0.07
0.25
212501_at
CEBPB
0.6
1.82
0.13
-1.31
-0.41
-0.73
-0.41
-3.12
-0.59
-0.8
221577_x_at
GDF15
2.93
0.4
0.13
-4.55
-3.49
-4.88
-0.68
-5.63
-0.77
-2
201162_at
IGFBP7
1.26
-0.27
0.34
-3.28
-1.06
-2.72
-1.08
-1.79
-1.36
-1.85
201163_s_at
IGFBP7
1.01
-0.27
0.27
-3.26
-0.83
-2.21
-0.84
-1.77
-1.05
-1.78
213910_at
IGFBP7
0.22
0.22
-0.14
-0.48
-0.11
-0.93
-0.23
-0.49
-0.36
-0.35
210118_s_at
IL1A
3.42
3.82
0.22
-4.36
-0.7
-3.28
-0.26
-5.18
-0.04
-1.05
208200_at
IL1A
0.04
0.41
-0.07
0.13
0.16
-0.29
0.26
0.74
0.33
0.12
205067_at
IL1B
3.33
3.54
0.06
-4.6
-1.41
-4.13
-1.48
-5.44
-0.86
-1.98
39402_at
IL1B
3.29
3.43
0.38
-4.58
-1.31
-3.83
-1.37
-5.89
-0.8
-1.82
203828_s_at
IL32
1.96
0.78
0.17
-1.42
-0.86
-1.23
-1.11
-0.26
-0.36
-0.99
205207_at
IL6
2.01
4.16
0.13
-3.37
-1.24
-2.65
-1.93
-5.35
-1.97
-2.64
211506_s_at
IL8
1.14
2.56
1.7
-5
-2.15
-5.77
-2.21
-7.47
-1.26
-2.03
202859_x_at
IL8
0.66
1.1
1.06
-3.26
-1.01
-4.13
-1.01
-5.89
-0.53
-1
218878_s_at
SIRT1
-1.2
0.52
-0.19
0.98
0.34
0.85
0.37
0.85
0.69
0.72
218065_s_at
TMEM9B
0.44
-0.27
-0.11
-0.43
-0.33
-0.29
-0.38
-0.13
-0.18
-0.2
222507_s_at
TMEM9B
0.39
-0.29
-0.19
-0.35
-0.28
-0.29
-0.34
0.11
-0.07
-0.12
201010_s_at
TXNIP*
2.45
0.5
0.24
-2.3
-0.41
-1.36
-0.89
-1.46
-1.78
-1.55
201008_s_at
TXNIP*
2.42
0.22
0.29
-2.51
-0.36
-1.58
-0.95
-1.45
-1.87
-1.54
201009_s_at
TXNIP*
1.54
0.13
0.27
-1.65
-0.35
-1.22
-0.71
-1.22
-1.39
-1.25
Table 4.6: Senescence specific changes in NF-κB target genes expression with complementation
NFKB targets after complementation with the indicated constructs are shown as well as the changes in
expression of all the NFKB targets examined in more details in this study (C).
194
However they were not affected by heat shock. Moreover up-regulation of these NF-B
targets was reversed when growth arrest was overcome by abrogating the p53-p21 or
p16-pRb pathways. In addition to IL1A and B and IL6, a number of other secreted
protein genes were found to be up-regulated including IL15, IL32, IL33, CCL2, CCL20,
CCL26, BMP2, GDF15, LIF, IGFBP4 and IGFBP5.
4.3.3 Is the NF-κB pathway also activated at a protein level?
To determine if the increase in RNA expression of the NF-B targets was associated
with increases in protein expression and secretion particularly of the SASP cytokines,
the levels of IL6 and IL8 were determined in 12 hour culture supernatants collected from
CL3EcoR cells at 34C and after a 3 week growth arrest at 38C; culture supernatants
from HMF3S cells growing at 34C and 38C were used as controls for the temperature
shift. Growth arrest was associated with a very large increase in the level of IL6 which
was not due to the temperature shift (Figure 4.12A). IL8 levels were also increased
upon growth arrest but not to the same extent as IL6 (Figure 4.12B).
Together these results show that in these conditionally immortal cells senescence growth
arrest results in altered expression of many targets of the NF-κB pathway and upregulation of a number of SASP proteins.
4.3.4 Is phosphorylation of RelA/p65 also induced?
Activation of NF-B signalling is habitually associated with increased phosphorylation
of RelA. To determine if RelA phosphorylation was increased upon growth arrest,
lysates prepared from CL3EcoR cells grown at 34C and after 7 days at 38C were
analysed by Western blot using in parallel an antibody specific for total RelA or RelA
phosphorylated on Serine 536. Protein extracts prepared from HMF3S cells grown at
34C and 38C were used as temperature controls.
195
A
B
Figure 4.12: Secretion of IL6 (A) and IL8 (B) by senescent cells
12 hour supernatants harvested from cells grown at 34°C or at 38°C for 21 days were analysed by
Quantiglo ELISAs from R&D Systems. All measurements are from independent biological triplicates.
196
Phosphorylation of RelA was increased upon growth arrest (Figure 4.13) in accordance
with the finding that the NF-B pathway was activated upon growth arrest.
4.3.5 What happens if the NF-κB complex is inactivated?
4.3.5.1 RNAi mediated silencing of NF-κB subunits
abrogates senescence growth arrest
Since the NF-B pathway was activated upon growth arrest of CL3 EcoR cells, I
determined whether this activation has a causative role by determining if growth arrest
would be overcome upon individual silencing of the various components of the NF-B
transcription factor complex. The silencing strategy chosen here was to use silencing
constructs from the Open Biosystems human GIPZ lentiviral shRNAmiR library.
Silencing constructs corresponding to RelA, RelB, NFKB1 and NFKB2 subunits were
individually transduced into CL3EcoR cells after packaging as lentiviruses in HEK cells.
Stably infected cells were selected in puromycin at 6g per ml, pooled and assayed for
complementation. Selection of the infected cells at 6 g per ml puromycin enriches for
the transduced cells that have the highest levels of shRNAmiR expression.
All
experiments were carried out in triplicate and numbers of densely growing colonies of
cells determined after 3 weeks at 38C.
Although none of the NF-B silencing constructs were as efficient as silencing
p21CIP1/WAF1/Sdi1, silencing of the NF-B subunits was clearly able to overcome growth
arrest (Figure 4.14). Some constructs yielded more colonies than others but at least two
constructs for each subunit yielded growing colonies. The numbers of growing colonies
obtained after silencing NF-κB components were very similar or slightly higher than
obtained upon inactivation of the p16-Rb pathway with HPV16 E7 or E2F-DB protein
and the growing phenotype of the obtained was very clear cut.
197
Figure 4.13: Increase in phosphorylation of RelA (Ser536) in senescent cells
Nuclear proteins extracted from cells grown at 34°C or 38°C for 12 days were analysed by western
blotting using Phospho- NF-κB p65 (Ser536) (93H1) and NF-κB p65 (C22B4; Cell Signalling).
198
Figure 4.14: Silencing of NF-κB transcription factor subunits
CL3EcoR cells were infected in duplicate with lentiviruses expressing the indicated GIPZ shRNAmir
silencing constructs. After puromycin selection, 0.5x105 stably transduced cells were seeded in triplicate,
incubated at 38°C for 21 days and stained. Densely growing colonies were counted after microscopic
examination.
199
4.3.6 Modulation of the NF-κB pathway overcomes senescence growth arrest
To confirm that the NF-B pathway has a causative role in senescence, it was modulated
both positively and negatively by RNAi mediated silencing and ectopic gene expression.
4.3.6.1 TMEM9B and
senescence
BCL2L1
silencing
bypass
In a parallel project, an RNA interference screen has been used to identify genes whose
suppression overcomes growth arrest of CL3EcoR cells (ER, LM and PSJ, manuscript in
preparation, see next chapter). One of the shRNAs isolated corresponded to TMEM9B.
TMEM9B was up-regulated 1.3 fold (P-value 1.47E-07, Table 4D) upon growth arrest
which was reversed when growth arrest was overcome. Since TMEM9B has been shown
to activate NF-κB dependent reporter constructs (Matsuda, Suzuki et al. 2003; Dodeller,
Gottar et al. 2008), silencing its expression should suppress the NF-κB pathway
resulting in abrogation of senescence growth arrest. Four lentiviral shmiRs targeting
TMEM9B from the Open Biosystems human genome wide GIPZ library were
introduced individually into CL3EcoR cells and the transduced cells analysed by the
growth complementation assay. Densely growing colonies were obtained after 3 weeks
growth at 38C (Figure 4.15). Another gene identified by the RNA interference screen
was BCL2L1, a member of the BCL2 family of proteins that is dependant and act
downstream of NF-κB. BCL2L1 was up-regulated 1.57 fold (P-value 9.24E-04, Table
4D) upon growth arrest which was reversed totally upon complementation with either
p53 or pRb abrogation. Silencing of BCL2L1 also overcame growth arrest (Figure 4.15)
with 3 lentiviral shmiRs constructs out of 3 tested.
200
Figure 4.15: Silencing of TMEM9B and BCL2L1
CL3EcoR cells were infected in duplicate with lentiviruses expressing the indicated pGIPZ shRNAmir
silencing constructs and assayed for growth complementation at 38°C. The number of growing colonies
were counted after 3 weeks.
201
4.3.6.2 Silencing of cEBPβ, BTG2 and TXNIP silencing
bypass senescence
Silencing of C/EBP a transcription factor, previously proposed to be up-regulated in
these conditional cells and linked to NF-B activity also abrogated growth arrest (Figure
4.16) with two silencing constructs out of two tested. Silencing of BTG2, an NF-κB
responsive gene (Kawakubo, Carey et al. 2004) and TXNIP, another member of the NFB pathway, as well as the secreted proteins CCL26, GDF15, IGFBP7 and IL32 were
also able to complement growth (Table 4.6C and Figure 4.17 and 4.18).
4.3.6.3 Ectopic expression of IKB-SR bypasses senescence
NF-B activity can also be suppressed by ectopic expression of a non-phosphorylatable,
dominant negative form of IB, the super-repressor of NF-B (IB-SR).
A
tetracycline inducible lentiviral expression construct for IB-SR was introduced into
CL3EcoR cells and the stably transduced cells assayed by the complementation assay
upon doxycycline induction. Expression of IB-SR overcame growth arrest further
indicating that activation of NF-B activity has a causative role in the growth arrest
(Figure 4.19).
202
Figure 4.16: Silencing of C/EBPβ
CL3EcoR cells were infected in duplicate with lentiviruses expressing the indicated pGIPZ shRNAmir
silencing constructs and assayed for growth complementation at 38°C. The number of growing colonie
was counted after 3 weeks.
203
Figure 4.17: Silencing of BTG2 and TXNIP
CL3EcoR cells were infected in duplicate with lentiviruses expressing the indicated pGIPZ shRNAmir
silencing constructs and assayed for growth complementation at 38°C. The number of growing colonie
was counted after 3 weeks.
204
Figure 4.18: Silencing of secreted proteins CCL26, IGFBP7, GDF15 and IL32
CL3EcoR cells were infected in duplicate with lentiviruses expressing a mix of shMIRS silecing constructs
for CCL26 (V2LHS_70279, V2LHS_70276 and V2LHS_70275 ) or the indicated pGIPZ shRNAmir
silencing constructs and assayed for growth complementation at 38°C. The number of growing colonie
was counted after 3 weeks.
205
Figure 4.19: Ectopic expression of IKB-SR
CL3EcoR cells were infected in duplicate with lentiviruses expressing Lamin A/C shRNA, pTIPz IkB-SR or
an empty pTIPz vector. Stably transduced cells were assayed for growth complementation at 38°C with or
without doxycycline. The number of growing was counted after 3 weeks.
206
4.3.6.4 Ectopic expression of SIRT1 bypasses senescence
To further confirm the role of NF-B activity, ectopic expression of SIRT1 was used to
suppress NF-B. SIRT1, the human homologue of Sir2, is a histone deactylase that has
been shown to suppress NF-B activity (Yeung, Hoberg et al. 2004). SIRT1 expression
was down-regulated 2.3 fold (P-value 9.22E-12; Table 4.6C) upon growth arrest which
was reversed upon complementation. Ectopic expression of SIRT1 promoted growth in
the complementation assay (Figure 2.20) and initial experiments suggest that this is
dependent upon the deacetylation function of SIRT1. However as SIRT1 also negatively
regulates p53 activity (Langley, Pearson et al. 2002), it may be overcoming growth
arrest by inactivating p53 rather than suppressing NF-B activity or acting on both.
4.3.7 NF-κB Activation is Causal to Senescence
Together the results show that the senescence growth arrest in these conditionally
immortalised fibroblasts involves activation of the NF-B pathway and that suppression
of this pathway by either direct silencing of NF-B subunits or by upstream modulation
can overcome growth arrest. Involvement of the NF-B pathway was further indicated
by silencing of BCL2L1, BTG2 and TXNIP that act downstream of NF-B and C/EBP
that may act in concert with NF-B to regulate gene expression.
207
Figure 4.20: Ectopic expression of SIRT1
CL3EcoR cells infected in duplicate with retroviruses pYESir2-puro and pLPCX were assayed for growth
complementation at 38°C.
208
4.4 DISCUSSION
One of the major stumbling blocks in dissecting the molecular pathways that underlie
cell senescence has been the asynchrony of this process in heterogeneous cell
populations used for serial sub-cultivation. The finding that HMF3A cells undergo an
essentially synchronous growth arrest within 7 days of shift up to 38 oC and that this can
be readily overcome has enabled us to combine genome wide expression profiling with
genetic complementation to identify genes that are differentially expressed when the
cells undergo proliferation arrest. This identified 961 genes which were down-regulated
>2 fold and 816 genes that were up-regulated >2 fold. Moreover when growth arrest
was abrogated by complementation, the differential expression was reversed; downregulated genes were up-regulated whereas up-regulated genes were suppressed.
Some of the genes identified, such as MAN1C1, PERP, DAB2, GM2A and PRNP have
previously been shown to be induced upon senescence (Wagner, Horn et al. 2008).
Many of the other up-regulated genes encode metalloproteinases and collagenases and
other extra-cellular matrix degrading enzymes that are involved in collagen turnover and
are hallmarks of a senescent cell microenvironment; for example, ADAMST1,
cadherin2, CD36, MT1F, MT1X, MMP10, MMP12 and TIMP1 (West, Pereira-Smith et
al. 1989; Millis, McCue et al. 1992; Yoon, Kim et al. 2004). Another large subset of upregulated genes included those that encode for secreted factors, including NRG1, FGF2,
VEGFC, CSF1, DAF, CD59, IL15, IL32, IL33, CCL2, CCL20, CCL26, BMP2, GDF15,
LIF, IGFBP4 and IGFBP5.
4.4.1 SASP: Senescence-Associated Secretory Phenotype
It has long been known within the field that the culture medium of senescent cells is
enriched with secreted proteins (Shelton, Chang et al. 1999; Krtolica and Campisi 2002).
The SASP concept was first proposed by the Campisi group, when they realized that
209
secreted factors from senescent fibroblasts promote the transformation of pre-malignant,
but not of normal, mammary epithelial cells. This initial observation of SASP implies
that senescence might not simply be a tumour suppressor mechanism, but rather a
double-edged sword within the tumour microenvironment. What remained unclear;
however, were the functional effects of SASP on the senescence phenotype itself. A
series of recent papers (Acosta, O'Loghlen et al. 2008; Coppe, Patil et al. 2008;
Kuilman, Michaloglou et al. 2008; Wajapeyee, Serra et al. 2008; Augert, Payre et al.
2009), have identified various new members involved in SASP and notably IL-6 and IL8 which are also up-regulated upon senescence in the HMF3A cells, and collectively
reinforced the idea that senescence is both regulated by and regulates the extracellular
environment.
4.4.2 Senescence Down-Regulated Genes
Many of the other down-regulated genes including those that are required for cell cycle
progression, cell proliferation and mitosis, were similarly identified in the original
HMF3A microarray study (Hardy, Mansfield et al. 2005), such as CDK4, CDC2,
CDC25B, CDC25C, BUBR1, PRC1, FOXM1 and UBE2C. Down-regulation of a
number of genes that encode proteosomal subunit components was also noted, namely
PSMB1, PSMB4, PSMB7, PSMB6, and this finding was in accordance with that of
Chondrogianni and colleagues (Chondrogianni, Stratford et al. 2003) who identified a
reduction in the expression level of catalytic subunits of the 20S proteosome and
subunits of the 19S regulatory complex upon the induction of senescence.
Many of the down-regulated genes are associated with the cell cycle and are generally
not affected by serum starvation (Table 4.2B). Three of the top four most highly downregulated transcripts (NUF2, SLC25 and NDC80) are all components of the NDC80
kinetochore complex.
The down-regulated transcripts also comprise genes that are
necessary for the transition from G1 to S phase (Cdc6 and Cdc25B) and G2 to M phase
(AurkB, Bub1 and Kif20a) as well as Polo like kinase 1 and MCM 4, 5 and 7. Many of
these are known to be direct targets of the E2F and FoxM1 transcription factors
suggesting that they are co-ordinately regulated by them. The association of the down210
regulated genes with the control of the cell cycle is consistent with the observed loss of
proliferative potential when cells undergo senescence. Moreover when HMF3A cells
cease dividing they undergo growth arrest in G1 phase in a manner analogous to
senescent cells that also predominantly arrest in G1 phase. Since no cells arrest in S
phase, it indicates that cells are unable to undergo the G1 to S transition consistent with
the finding that genes involved in the G1 to S transition were down-regulated. The
down-regulation of genes related to G2 phase is particularly interesting because DNA
synthesis can be induced in senescent cells by exposure to fresh mitogens or by superinfection with DNA tumour viruses but the cells will not undergo mitosis suggesting
there may also be block in the G2 phase in senescence (Gorman and Cristofalo 1985).
Using a rodent model of senescence we have previously proposed that senescence
involves a growth arrest in both G1 and G2 and that the block in G2 may actually be the
cause of the irreversible loss of proliferative potential (Gonos, Burns et al. 1996).
4.4.3 Ectopic expression of down regulated genes rescues the growth arrest
Out of six down-regulated genes chosen to be tested, two permitted rescue in a
significant manner, namely BUB1B and FOXM1, when the other four did not give
convincing rescue. However, since the failing constructs expression was not confirmed
by western blot analysis, it is not possible to draw conclusion on their actual efficiency
on bypassing senescence. Another three genes did not yield sufficient number of puro
resistant clones and therefore could not be tested appropriately.
We have noted previously that in the HMF3A complementation assay, the density of the
cells was an essential parameter to keep in consideration as a low density could cause a
stress which would not permit the rescue from senescence.
HMGB2 and NEK2 did not yield growing colonies; however, their expression was not
confirmed and therefore their efficiency to rescue cannot be assessed.
211
DEPDC1 was able to bypass cellular senescence at a very low level but with very
variable results and therefore was not considered significant enough. It was previously
reported that silencing of DEPDC1 was able to inhibit growth of bladder cancer cells
(Kanehira, Harada et al. 2007), therefore it seemed possiblel that its expression would
permit the growth in the conditional fibroblasts. This should be investigated further in
the future.
BUB1 was a potential target to test in our system as it was believed to function as a
growth enhancer protein in several other models (Cahill, Lengauer et al. 1998; Gjoerup,
Wu et al. 2007; Gao, Ponte et al. 2009). It was then expected that its expression in the
CL3EcoR cells would bypass senescence. Its ectopic expression permitted the bypass of
the growth by yielding healthily growing colonies at a level above background.
MELK was one of the targets, that I thought most promising in the context of breast
cancer and brain cancer as its expression was found to be directly linked to the
malignancy of several cancers including breast cancer and as its silencing was also
linked to growth suppression (Lin, Park et al. 2007; Marie, Okamoto et al. 2008;
Nakano, Masterman-Smith et al. 2008). However, the results did not show a bypass of
senescence. Moreover, Catia Caetano blotted the cells with a MELK antibody and could
not detect any protein expression so no comment about MELK ectopic expression effect
on senescence can be made. It would be worth investigating MELK further perhaps
with a new expression construct.
MLF1-IP was a potential target to test in this context as its expression has been
previously linked to several cancers malignancy and the escape from senescence in
murine embryonic fibroblasts (Matsumoto, Yoneda-Kato et al. 2000; Dornan, Wertz et
al. 2004; Sun, Zhang et al. 2004; Hanissian, Teng et al. 2005; Yoneda-Kato, Tomoda et
al. 2005; Yoneda-Kato and Kato 2008). However, the level of rescue was considered too
low to be of interest. In addition, the expression of MLF-IP was not confirmed and
therefore no conclusion could be made about its efficiency to bypass senescence.
212
Transcriptionally active FOXM1 was sufficient to bypass senescence and was shown to
have a causative role in cellular senescence but further studies are required to confirm
and extend these findings to which isoform(s) are differentially expressed, how they are
activated, what causes their expression to be down-regulated and what is their
mechanism of action.
Indeed, I have found that senescence in CL3 EcoR cells can be abrogated by the
constitutively active FOXM1c but not the wild type protein. This indicates a requirement
for the actual activation of FOXM1. The N-terminus of FOXM1 contains an autorepressor domain that inhibits transactivation by an intramolecular interaction with the
C-terminal TAD (Figure 4.8A). This repression can be relieved by phosphorylation of
multiple cdk sites within the TAD by cyclinA/cdk2 or possibly by cyclinE/cdk2; PLK1
and PLK4 may also play a role. The expression profiling data indicates that cdk2 and
cyclinE expression are unaffected upon growth arrest whereas cyclinA expression is
down-regulated about 20 fold, PLK1 30 fold and PLK4 12 fold respectively.
Unfortunately, when PLK4 was ectopically expressed to be tested here, it did not yield a
sufficient number of puro-resistant cells, and therefore could not be tested. Maybe a
different expression system should be used.
Another important question would be to determine what causes the down-regulation of
FOXM1 upon cell senescence. Although it was recently suggested that that Stressactivated kinase p38 (p38SAPK) is capable of inhibiting FoxM1 expression (Adam et al,
2009), the transcription profiling data indicates that this unlikely to be the mechanism,
since expression of the three isoforms α (MAPK14), β (MAPK11) and γ (MAPK13) of
p38SAPK present in HMF3A cells, is unaffected upon growth arrest.
Previously it was suggested that in Basal Cell Carcinomas, FOXM1 was a downstream
target of Gli1, which is transcriptionally up-regulated by Sonic hedgehog (Shh)signalling (Teh, Wong et al. 2002). Gli1 is a member of the Gli family of three
transcription factors Gli1, 2 and 3. Gli 1 and 2 are activators whereas Gli3 is a repressor.
213
The expression profiling data shows that all three Gli proteins are expressed in
proliferating CL3EcoR cells but upon growth arrest Gli2 and 3 are down-regulated
whereas Gli1 may be slightly up-regulated. This could be another explanation for the
down-regulation of FOXM1 upon senescence.
In conclusion, many of the down-regulated genes are associated with the cell cycle
(Table 4.1B and Supplementary Table S4.2). They comprise genes necessary for the
transition from G1 to S phase (CDC6 and E2F1), G2 phase (CDC22, TOP2A and
cyclinA) and G2 to M phase (AURKB, BUB1 and KIF20A) as well as Polo like kinase
1 and MCM 4, 5 and 7 (Whitfield, Sherlock et al. 2002). Many of these are direct targets
of E2F and FOXM1 suggesting they are likely to be co-ordinately regulated by them.
This is supported by my finding that constitutively active FOXM1 abrogated
senescence. The association of these down-regulated genes with cell cycle control is
consistent with the loss of proliferative potential when cells undergo senescence growth
arrest mostly in G1. However, the down-regulation of genes related to G2 is interesting
because DNA synthesis can be induced in senescent cells by exposure to fresh mitogens
or by super-infection with DNA tumour viruses but the cells do not undergo mitosis
suggesting there may also be block in G2 in senescence (Gorman and Cristofalo 1985).
Using a rodent model of senescence, Pr Parmjit Jat had previously proposed that
senescence involves growth arrest in both G1 and G2 and that the block in G2 may
actually be the cause of the irreversible loss of proliferative potential (Gonos, Burns et
al. 1996).
4.4.4 NF-κB Pathway Activation upon Senescence is causal to Senescence
An in silico analysis of our differential data set and previous data suggested a potential
activation of NF-κB upon senescence (Hardy, Mansfield et al. 2005). I investigated
further the involvement of NF-κB pathway in senescence and used different approaches
to validate in vitro my initial hypothesis.
214
4.4.4.1 Senescence Up-Regulated Genes
In contrast to the down-regulated genes which are unaffected by serum starvation, many
of the up-regulated genes were significantly up-regulated by serum starvation (IL33,
ABI3BP, IL1A, IL1B and PAPPA amongst the top 24 up-regulated genes). Four of the
top five most highly up-regulated transcripts correspond to CLCA family member 2,
chloride channel regulator. The Chloride Channel Accessory (CLCA) family of proteins
has four members in humans, are widely expressed in variety of cell types and encode
900 amino acid proteins that have been shown to produce a novel Cl- current that can be
activated by calcium and inhibited by Cl- channel inhibitors. Elble and colleagues have
recently shown that hCLCA2 is a direct downstream target of p53 and its acute
expression results in a senescence like cell cycle arrest or apoptosis depending upon the
cell context (Walia, Ding et al. 2009); Elble, personal communication).
The up-regulated genes also comprises a number of secreted proteins such as IL1A and
B, IL6, IL15, IL32, IL33, CCL2, CCL20, CCL26, BMP2, GDF15, LIF, IGFBP4 and
IGFBP5; IL8 and IGFBP7 were also up-regulated but since the adjusted P-values were
greater than 0.0001 they were not identified as significant. IL6 and IL8 were also shown
to be secreted into the medium. IL6, IL8 and IGFBP7 have been found to be secreted by
senescent cells and act cell autonomously to induce and reinforce senescence (Acosta,
O'Loghlen et al. 2008; Coppe, Patil et al. 2008; Kuilman, Michaloglou et al. 2008;
Wajapeyee, Serra et al. 2008; Adams 2009; Augert, Payre et al. 2009; Kuilman and
Peeper 2009; Orjalo, Bhaumik et al. 2009). Kuilman et al (2008) have further suggested
that induction of these SASP proteins was specific to oncogene-induced senescence and
not affected by quiescence. The results (Table 6C) indicate that these secreted proteins
are also strongly up-regulated by serum starvation. Acosta et al (2008) have shown that
activation of these secreted chemokines was regulated by the NF-κB and C/EBPb
transcription factors; the results discussed below are in accordance with these findings.
Orjalo et al (2009) have suggested that IL1A, one of the most highly up-regulated genes
was an essential cell-autonomous regulator of the senescence–associated IL6/8 cytokine
network. In addition to IL6, IL8 and IGFBP7, I have found that a number of other
secretory proteins were up-regulated. Moreover silencing expression of CCL26, IL32
215
and GDF15 was capable of abrogating growth arrest indicating that they may also have a
role in promoting and reinforcing senescence. However, it remains to be demonstrated
whether CCL26, GDF15 and IL32 play a similar role in promoting senescence in
primary fibroblasts and whether the other up-regulated cytokines have a similar role.
4.4.4.2 Silencing of over-expressed genes bypasses the
growth arrest
Out of nine genes up-regulated upon senescence chosen to be tested, six permitted the
bypass of the conditional growth arrest, two gave a weak rescue and one did not rescue
at all. This highlights the importance of mRNA expression regulation in the senescence
mechanisms.
AK3L1, CLCA2, SCN2A, GRAMD3, TRIB2, p16INK4A, BLCAP and RUNX1 silencing
all permitted the rescue of the cells from senescence using a mix of silencing constructs
or individual silencing constructs. Interestingly, rescue by silencing of TRIB2 is
contradictory to the literature as it was reported to be able to inactivate cEBPalpha and
beta (Naiki, Saijou et al. 2007) which are not only known to be involved in senescence.
In addition, cEBPbeta silencing permitted rescue in the HMF3A cells.
CDKN2A or p16INK4A silencing also permitted to bypass the growth arrest; this further
reinforces the hypothesis that abrogation of the pRb alone was sufficient to bypass the
conditional growth arrest.
RUNX1 silencing did permit the bypass of senescence in the HMF3A system. This was
in accordance to several studies where RUNX1 was able to induce senescence-like
growth arrest in primary murine fibroblasts (Linggi, Muller-Tidow et al. 2002; Wotton,
Blyth et al. 2004; Kilbey, Blyth et al. 2007). This senescence-inducing effect is spared in
cells lacking expression of p19 Arf, an inducer of the p53 pathway (Linggi, Muller-Tidow
et al. 2002). More recently another study suggested that this senescence inducing effect
was actually happening through pathways independent to the one of replicative
216
senescence. Indeed, RUNX1 induced early growth stasis with only low levels of DNA
damage signalling and a lack of chromatin condensation, a normal marker of irreversible
growth arrest. In human fibroblasts, RUNX1 also induced p53 in the absence of
detectable p14ARF but not in the absence of p16INK4A (Wolyniec, Wotton et al. 2009).
BLCAP silencing did not permit rescue in the HMF3A model even if its ectopic
expression was linked to growth inhibition in several cell line and models (Zuo, Zhao et
al. 2006; Yao, Duan et al. 2007).
GRAMD3 permitted the bypass of senescence and represent a totally novel target to
study in the context of the cell cycle and senescence. In addition, the fact that its
expression seems regulated by two of the miRs whose expression can overcome
senescence in chapter 4 makes it an even more credible target in the senescence
pathway.
SCN2A and CLCA2 silencing both showed rescue at a low level and each of the repeat
experiment showed the same results. This suggests a role, even minor, for these two
genes both involved in ion conductance. Kriete and Mayo (2009) have proposed that
mobilisation of calcium stored within the endoplasmic reticulum in conjunction with
increases in reactive oxygen species from the mitochondria activates NF-κB signalling.
The involvement of calcium in senescence is intriguing since the most highly upregulated gene I identified was CLCA2. CLCA2, a chloride channel regulator was also
found to be able to induce a senescence-like growth arrest upon acute expression (Walia
et al, 2009; Elble, personal communication).
4.4.4.3 Senescence expression profile reveals links with
Cancer expression profile
Overlay of the differential data set with the meta-signatures of genes that are upregulated upon neoplastic transformation or undifferentiated cancer (Rhodes, Yu et al.
2004) showed that nearly 50% of these genes were down-regulated when HMF3A cells
undergo growth arrest. Hanahan and Weinberg (2000) have proposed that six
217
independent events are required for malignant transformation and the acquisition of an
infinite proliferative potential is one of these events. The results indicate that even
though overcoming senescence may only be one of the six events, 50% of the genes are
related to it and thus it must be an important barrier in cancer development.
4.4.4.4 NF-κB pathway is activated upon senescence
Our previous study suggested that the loss of proliferative potential in these
conditionally immortal fibroblasts may involve activation of the NF-κB pathway (Hardy,
Mansfield et al. 2005). I therefore analysed the promoters of the differentially expressed
genes and found that NF-κB motifs were amongst the top 10% of most abundant motifs
in both the up- and down-regulated genes.
A role for NF-κB activation was further highlighted by the finding that 67 NF-κB targets
were up-regulated against 26 that were down-regulated significantly upon growth arrest
(Table 4B and C); this included the increased expression of SASP proteins, including
IL6, IL8 and IGFBP7 that are known to promote and reinforce senescence. Expression
of other NF-κB targets was also differential but since the P-values were greater than
0.00001, they were not considered to be significant. In contrast to the down-regulated
genes almost all of the up-regulated NF-κB targets were also up-regulated upon serum
starvation.
Moreover, the increased expression, including the SASP proteins was reversed upon
abrogation of the growth arrest by almost all constructs; even though expression of the
down-regulated genes was also reversed it was not as clear cut. Although NF-κB is
generally suggested to promote growth, my finding that it may be associated with
growth arrest is in accordance with the study of Penzo et al (2008) who have shown that
acute activation of NF-κB in murine embryo fibroblasts results in growth arrest in
association with repression of 20 genes essential for cell cycle progression that are
known targets of either E2F or FOXM1. Comparison of these genes with the differential
GA set indicates that all of these genes except CDC5A were significantly repressed upon
growth arrest (Supplementary Table S4.2). FOXM1 and E2F were down-regulated 6.23
218
and 1.8 fold respectively upon senescence growth arrest and the down-regulation was
reversed when senescence was overcome (Supplementary Table S4.2).
Moreover
constitutively active FOXM1 (NKEN) and E2F-DB overcome senescence growth
arrest suggesting that transcriptionally active forms of these transcription factors may
have a causative role in senescence.
 Diversity of the NF-κB function
The NFB family of ubiquitously transcription factors are widely known as key
regulators of inflammatory and immune response. More recently they have been shown
to function as regulators of diverse cellular processes such as cell proliferation and
differentiation and the response to stresses such as oxidative, physical and chemical
stress. Activation of NF-B also blocks apoptosis and promotes cell survival. This
family of transcription factors consists of five proteins, RelA (p65), RelB, c-Rel,
NFKB1 (p105/p52) and NFKB2 (p100/p52) that are related through the Rel homology
(RH) domain, a highly conserved DNA-binding and dimerisation domain.
They
associate to form homo- and heterodimers and regulate transcription by binding to B
motifs within flanking DNA sequences.
In unstimulated cells, the complexes are
retained within the cytoplasm by binding to a member of the IB family of inhibitory
proteins that bind to and mask the nuclear transport sequence. Upon stimulation, the
IB protein is phosphorylated resulting in its ubiquitination and degradation leading to
liberation of the NF-B complex, translocation to the nucleus and binding to target
DNA. The IB proteins are phosphorylated by the IB kinases which consist of three
subunits: the catalytic subunits IKK and IKK and the regulatory subunit IKK
(NEMO).
Each of the above components is integrated into a complex signalling
network central to the control of NFB activity.
 NF-κB and ageing
Since my initial finding that NFB activity may be associated with senescence, Adler
et al (2007) using a systematic bioinformatic approach to identify combinatorial cis219
regulatory motifs showed that NFB activity controlled cell cycle exit and was
continually required to enforce many features of ageing in a tissue specific manner.
Moreover activation of NFB with age is consistent with elevated levels inflammatory
markers and a pro-inflammatory phenotype associated with many age related diseases.
For instance, factors that mediate NFB and inflammation include the insulin/IGF
pathway, SIRT1, FOXO, PDC-1 and PPAR. NF-B is also constitutively activated in
older human subjects compared to young donors (Kriete, Mayo et al. 2008). Moreover,
RelA has been proposed to maintain cellular senescence by promoting DNA repair and
genomic stability (Wang, Jacob et al. 2009).
The recent findings that senescence is associated with secretion of SASP proteins
including the inflammatory cytokines IL6 and IL8, further suggest a role for NFB
activation for inducing and reinforcing senescence.
 NF-κB Causal Role in Senescence
Our study shows that NFB has a direct causal role in senescence. The expression
profiling results in Table 4.6C show that the NFB transcription factor subunits are
themselves not differentially expressed upon senescence growth arrest. Their expression
was also not consistently modulated upon abrogation of senescence by the different
constructs. However, I have shown that silencing of NFB transcription factor subunits
by RNA interference clearly abrogated growth arrest upon activation of the p16-pRb and
p53-p21 pathways (Figure 4.7).
In addition, my RNA interference growth promotion screen has identified TMEM9B a
known upstream activator of NFB (ER, LM and PSJ, manuscript in preparation, see
Chapter 5). Here, I show that TMEM9B was up-regulated upon growth arrest (Table
4.6) and that shRNA mediated silencing of TMEM9B can overcome growth arrest in the
CL3EcoR cells (Figure 4.15). Silencing of BCL2L1, a direct downstream mediator of
NFB and C/EBP, a transcription factor linked to NFB also overcomes growth
220
arrest. Moreover constitutive expression of IB-SR, the super-repressor of NFB, also
abrogated growth arrest suggesting that NFB activation is likely to be via the
canonical pathway. Activation of the canonical NF-B pathway was also indicated by
the increased phosphorylation of RelA-Serine 536 upon growth arrest (Figure 4.13).
Kriete and Mayo (2009) have proposed another potential mechanism for NFB
activation in ageing, the atypical pathway. They proposed that mobilisation of calcium
stored within the endoplasmic reticulum in conjunction with increases in reactive
oxygen species from the mitochondria activates NFB signalling. The involvement of
calcium in senescence is intriguing since the most highly up-regulated gene I identified
was CLCA2, a chloride channel regulator and CLCA2 can induce senescence like
growth arrest upon acute expression (Walia, Ding et al. 2009); Elble, personal
communication).
Our study has clearly delineated a central role for NF-κB activity in cellular senescence
irrespective of whether it is induced via the p53-p21 or the p16-pRb tumour suppressor
pathway. It indicates that senescence growth arrest is associated with activation of the
canonical signalling pathway resulting in up- and down-regulation of known target
genes including the SASP cytokines IL6 and IL8 that can act cell autonomously to
induce and reinforce senescence. Moreover, this activation could also down-regulate
FOXM1 and E2F and their downstream targets that are critical for cell cycle progression
particularly the G2 phase. Although it is not clear how activation of the p53-p21 or p16pRb pathways results in activation of NF-κB signalling, one possibility is SIRT1, a
deactylase that can suppress NF-κB activity. SIRT1 expression was down-regulated
upon senescence but reversed upon its abrogation. Another possibility is TMEM9B, a
lysosomal transmembrane protein that can activate NF-κB dependent expression;
TMEM9B was slightly up-regulated upon senescence and reversed when it was
overcome. Although it remains to be demonstrated whether SIRT1 and TMEM9B are
the signal and what causes their differential expression, this study shows that NF-κB
activation has a causal role in promoting senescence and suggests a framework for
dissecting the underlying signalling network.
221
5
AN RNA INTERFERENCE SCREEN IDENTIFIES DOWNSTREAM
EFFECTORS OF THE P53-P21 AND P16-PRB PATHWAYS
5.1 RNAI INTERFERENCE SCREEN
5.1.1 Objectives
The discovery of RNAi has revolutionized the way investigators approach the studies of
gene expression, regulation and interactions, particularly as it relates to drug
development. It is a powerful tool which has been widely utilised in a variety of cells
lines to perform loss-of-function genetic screens and identify target genes involved in
various cellular processes (Zender, Xue et al. 2008; Hu, Kim et al. 2009; Zhang, Binari
et al. 2010). Here, the objective was to silence by RNAi 9,392 different cancer
associated genes in the conditional HMF3A cells and see if that would abrogate
senescence growth arrest by using a retroviral shRNA (short hairpin RNA) library
specifically designed for application in mammalian systems (Berns, Hijmans et al. 2004;
Paddison 2008).
I have shown previously that inactivation of the p53-p21, the p16-pRb or the NF-κB
pathway individually were able to bypass growth arrest in this model but the signal
transduction pathways involved and how these diverse signals that result in senescence
are all integrated, remain poorly defined. These functional assays would allow the
identification of any genes whose silencing overcome senescence growth arrest and
therefore any genes involved in and causal to the growth arrest. Potential positives were
confirmed by carrying out a secondary screen using either pools of lentiviral shmirs or
individual lentiviral shmirs.
222
5.1.2 The Open Access RNAi project at UCL
RNA interference is a natural system that exists within living cells to control and
regulate the levels of expression of genes at the mRNA level. There are two types of
small RNA molecules which are miRNAs (micro-rnas) and siRNAs (small interfering
RNA) utilized here as shRNA (short hairpin RNA).
The RNAi library that I used was originally developed by Greg Hannon (CSHL) and
Steve Elledge (Harvard) as a retroviral library. I used the pSM2 library version 1.3
provided by Chris Lord and Alan Ashworth (Breakthrough breast cancer research
center). This version was extended later to make it genome-wide and also cloned into the
pGIPZ lentiviral shRNAmir vector which I used to validate the results from the first
version. Clones forming this library were provided by the UCL RNAi consortium.
―Functional genomics is playing an ever important role in deciphering the roles of
specific genes in cancer and developmental biology, as well as in neuro-sciences,
infectious diseases and immunity. The Open Access RNAi program helped foster this
collaboration between the UCL Cancer Institute, UCL Institutes for Child Health and
Neurology, and Division of Infection and Immunity, enabling us to provide world-class
scientists in central London access to the latest shRNA libraries for focused functional
screens.‖ said Dr. Chris Boshoff, Director, UCL Cancer Institute.
5.1.3 Which viral delivery system? Which library?
The Open Biosystems Expression ArrestTM pSM2 retroviral shRNAmir Library is a
whole genome RNAi resource for transient, stable and in vivo RNAi studies. The
collection has several unique features that make it a very versatile and efficient tool for
RNAi studies including large-scale screens (Paddison, Silva et al. 2004) and notably its
human micro-rna-30 (miR30) adapted design (Figure 5.1) which increases knockdown
specificity and efficiency (Boden, Pusch et al. 2004).
223
Figure 5.1: miR-30 adapted shRNAmirRtranscript design
Expression Arrest shRNA are expressed as miR30 primary transcripts (figure from OpenBiosystems
literature).
224
The pSM2 retroviral shNA human library version 1.3 consisted of 100 pools numbered 1
to 100. The 100 tubes of plasmids pools each contains between 150 to 200 different
shRNA constructs using the pSM2 plasmid (Figure 5.2). Each gene is represented by 1
to 3 shRNA plasmids and each plasmid is complementary to a different region of the
target gene. Multiple shRNA plasmids per gene are used in order to increase the
likelihood of achieving maximum efficiency of gene knockdown. The library version 1.3
contained 15,148 constructs representing 9,392 human genes targeted.
The fact that only 9,392 genes were analysed in this experiment with respect to the
22,500 estimated number of unique human genes (International Human Genome
Sequencing Consortium, 2004) has to be taken into consideration. However, these 9392
genes were enriched for cancer-associated genes which make them very relevant to
studying cell cycle disruption. This makes this screen a good tool to indentify new
targets involved in senescence.
The pGIPZ was developed in a similar manner combining the design advantages of
micro-rna-adapted shRNA (shRNAmir) with the pGIPZ lentiviral vector (Figure 5.3) to
create a powerful RNAi trigger capable of producing RNAi in most cell types including
primary and non-dividing cells. This library gives access to 44,602 lentiviral silencing
constructs corresponding to 18,076 different human genes. Here, constructs from this
library were used to validate the results of the primary screen by silencing genes
individually rather than in pools.
225
Figure 5.2: pSM2 retroviral plasmid : design (A) and features (B)
(figure from OpenBiosystems literature).
226
Figure 5.3: pGIPZ lentiviral plasmid : design (A) and features (B)
(figure from OpenBiosystems literature).
227
5.1.4 ShRNA Screening Strategy
The screen was performed on the HMF3A conditional cells particularly using the
CL3EcoR cells. The cells were infected with each of the 100 pools and reseeded (to avoid
the density bias on cell growth) before being shifted to 38°C for 3 weeks. Because these
cells have the particular properties to grow at 34°C but arrest at 38 °C when
thermolabile LT is inactivated (Figure 5.4), the flasks showing growing colonies after 3
weeks of cultures would be considered to contain the candidate shmirs of the primary
screen (Figure 5.4).
The genomic DNA would then be extracted for these growing cells to identify the
proviral shRNA inserts responsible for the growth arrest bypass and the corresponding
involved gene. As it was difficult to ring clone colonies near each other and because
there was very few colonies in some flasks, the total growing cultures were trypsinised
and re-plated to enrich for growing cells. Owing to the presence of multiple inserts in the
whole cultures and in order to identify each individual shRNA construct, an extra step of
bacterial cloning was added before sequencing the inserts. All hits should have been
analysed in a secondary screen to eliminate false positives; however, here, due to the
limiting time scale, the microarray expression profiling data and literature were used to
prioritise the order in which they were tested.
5.1.5 Sensitivity of the model
Before starting the screen, it was important to get a proof of feasibility of an assay
combining our HMF3A system and the shRNA pool format. It was also important to
determine the sensitivity of the assay. Each of the pools contains between 150 and 200
shRNA constructs; consequently, it was necessary to test that a mix containing a positive
control construct diluted 1/200 in a negative control was able to trigger a rescue and this
was identifiable.
228
Figure 5.4: ShRNA screen strategy
The screening was performed on the conditional CL3EcoR cells. These cells have the particular properties
to grow at 34°C (A) but arrest at 38 °C (B) when thermolabile LT is inactivated . The cells were infected
with each of the 100 pools and reseeded before being shifted to 38°C for 3 weeks (C) ; The flasks showing
growing colonies after 3 weeks of cultures would be considered containing the candidate shmirs of the
primary screen.
229
Therefore, a mixture was created by mixing a quantity of positive previously tested pRS
p21F RNAi construct at 1/200 with negative pRS Lamin A/C construct. This spiked mix
was packaged as described in Material and Method in Phoenix Ecotropic cells and used
to infect CL3EcoR at 0.5x106 in a T-75 cm2 flask. Along with it, a positive control, p21F
RNAi and a negative, Lamin A/C construct were each packaged and used to infect a
flask of cells as above. After selection, the cells were trypsinised and reseeded at 8.5x10 4
per 15cm plate or 0.5x104 per well in 6-well plates. The next day the media was changed
and cells were shifted to 38°C for 3 weeks. At that point, the cells were stained using
methylene blue dye.
The Lamin negative control (Figure 5.5, A and D) showed very little background
whereas the p21 positive control (Figure 5.5, B and E) produced a confluent monolayer.
The spiked mixture Lamin/p21 exhibited numerous distinguishable growing colonies.
These results suggest that the sensitivity of the test was sufficient to permit the
identification of 1 construct in a mix of 200; therefore, the proof of concept of the model
was verified. This test was performed with a batch of CL3 EcoR from the passage p22+9
to make sure that cells from the exact same batch could be used for the screen itself.
5.1.6 Confidence intervals
Using the formula: ln (1-.95) / ln (1-1/(Library Size)) which is recommended for
genetic screens (see http://www.stanford.edu/group/nolan/screens/screens.html) by the
Nolan lab, it is possible to calculate the number of infectious events that needed to be
screened depending on the size of the library and the confidence interval required (Table
5.1). In that case, for a confidence interval of 99%, the number of infection events for
the screening of our library of 9393 genes would have to be superior or equal to 43,254.
This suggested that for each pool approximately 1000 independent infectious events
would be sufficient for a 99% confidence that all shRNAs within a pool had been
sampled.
230
Figure 5.5: Screen sensitivity test
Cells were infected with either pRS Lamin A/C (A and D), either pRS p21F RNAi (B and E) or a mix
1/200 p21F/lamin (C and F). After puromycin selection, cells have been reseeded at 8.5x10^4 per 15 cm
plate (A, B and C) or 0.5x10^4 per well in 6-wells plates (D, E and F) and shifted to 38°C for 3 weeks.
The plates were then stained and photographed.
Confidence
Interval (%)
9393
Genes
200
Genes
95%
28137
598
98%
36744
780
99%
43254
919
Table 5.1: shRNA screen confidence intervals
The table displays the number of infectious events required to be screened depending on the size of the
library and the confidence interval.
231
In the shRNA screen, the number of cells reseeded after puromycin selection was
5,3x104 which is really superior to 919 so the confidence in the results are superior or
equal to 99%. However, to ensure that the screen would be saturating, virus sufficient to
give rise to 10,000 infectious events was utilised for each pool (Table 5.2).
5.1.7 Titration of Phoenix Eco viral Supernatants
CL3EcoR cells are very susceptible to cellular stress under low cell density growth
conditions. As such, an experiment was performed to determine the quantity of titrated
viral supernatant required to obtain approximately 10,000 infectious events for each
pool.
Cells were seeded at 6x104 cells per well in 6-well plates (day 0) and infected the next
day (day 1) with different volumes (from 0.5ml to 1x10 -4 ml) of each virus pool in
presence of 8μg/ml polybrene. After 2 weeks puromycin selection at 34°C, the cells
were stained with methylene blue and the number of colonies counted. The volume of
virus required to produce approximately 10,000 infectious events was calculated for
each pool (Table 5.2). Unfortunately, because the amount of DNA available to us was
limited, for the pools with a low titer, the volume of viral supernatant used was set at 10
ml (maximum amount harvested).
5.1.8 Primary screen in the HMF3A: Procedure
The screen was performed in successive waves of 10 pools. Cells were seeded at 0.5x10 6
per T-75 cm2 flask on day 0 and infected on day 1 with the determined volume of virus
supernatant (Table 5.2) in presence of 8μg/ml polybrene. Each time, a positive control,
pRS-p21F RNAi construct virus supernatant, and a negative control, pRS-Lamin A/C
construct virus supernatant, were each used to infect a flask of cells.
232
volume
volume
volume
volume
DNA Pool virus used DNA Pool virus used DNA Pool virus used DNA Pool virus used
(ml)
(ml)
(ml)
(ml)
1
9.0
25
10.0
51
10.0
75
10.0
2
9.0
26
10.0
52
10.0
76
10.0
p21
5.0
27
10.0
p21
1.3
77
10.0
lamin
5.0
28
10.0
lamin
1.3
78
10.0
3
8.0
29
10.0
53
10.0
79
10.0
4
8.0
30
10.0
54
8.7
80
10.0
5
8.0
31
10.0
55
10.0
81
10.0
6
2.0
32
10.0
56
10.0
82
10.0
7
8.0
p21
1.3
57
10.0
p21
5.0
8
8.0
lamin
1.3
58
6.8
lamin
5.0
9
8.0
33
10.0
59
7.7
83
10.0
10
8.0
34
10.0
60
8.6
84
10.0
11
8.0
35
10.0
61
9.6
85
10.0
12
8.0
36
10.0
62
10.0
86
10.0
p21
1.3
37
10.0
p21
1.3
87
10.0
lamin
1.3
38
10.0
lamin
1.3
88
10.0
13
10.0
39
10.0
63
10.0
89
10.0
14
10.0
40
10.0
64
10.0
90
10.0
15
10.0
41
10.0
65
10.0
91
10.0
16
10.0
42
10.0
66
10.0
92
10.0
17
10.0
p21
1.3
67
10.0
p21
5.0
18
10.0
lamin
1.3
68
10.0
lamin
5.0
19
10.0
43
10.0
69
10.0
93
10.0
20
10.0
44
10.0
70
10.0
94
10.0
21
10.0
45
10.0
71
10.0
95
10.0
22
10.0
46
10.0
72
10.0
96
10.0
p21
1.3
47
10.0
p21
5.0
97
10.0
lamin
1.3
48
10.0
lamin
5.0
98
10.0
23
10.0
49
10.0
73
10.0
99
10.0
24
10.0
50
10.0
74
10.0
100
10.0
Table 5.2: Virus pools titration and supernatant volume used
233
The negative control permitted the evaluation of the level of background for each
experiment and the positive control made sure that the complementation assay worked in
these conditions. The media was changed on day 2 and puromycin selection at 2µg/ml
was added on day 4. On day 8, the cells were trypsinised, counted and reseeded at
5.3x104 per T-75cm2 flask or 1.8x104 per T-25cm2 flask in as many flasks as possible in
order to screen a representative number of cells. The next day the media was changed
and cells were shifted to 38°C for 3 weeks. After 3 weeks, multiple growing colonies
were observed in the p21 shRNA-infected CL3EcoR cultures; however, minimal
background was observed in the Lamin A/C shRNA-infected CL3EcoR cultures. The
number of colonies for each pool was counted and each colony was checked under the
microscope for a growing phenotype. The number of cells counted just before reseeding,
the number of flasks reseeded and the number of colonies for each flask reseeded after 3
weeks at 38°C is shown in Table 5.3. The hits that were significantly above background
(more and/or bigger colonies, all analysed for phenotype under the microscope) were
reseeded (and are shown in red in Table 5.3). Occasionally, a background of growing
colonies was observed. For these batches, only the flasks showing a growth above
background were selected. The results show that 34 pools out of 100 yielded growing
colonies at a level above background. Particularly, the pools 13, 78 and 82 gave colonies
in a higher number and most importantly of a size superior to that of the other hits. Pools
16, 18, 19, 21 and 80 also yielded colonies but they were smaller. For each pool that
contained growing colonies, 1 to 4 flasks containing the highest number of colonies
were reseeded for extracting genomic DNA and resulted in a total of 81 sub-pools to
analyse. Genomic DNA was extracted from 80% confluent T-75 cm2 cultures. As a
caveat, it should be noted that genomic DNA harvested from these cultures were
representative of a heterogeneous population from multiples colonies and of many
different shRNAs species, therefore, sequencing of multiple inserts was required to
identify all shRNA target sequences. Nonetheless, a secondary growth complementation
screen of CL3EcoR cells utilising the pure shRNA construct would permit identification
of functional inserts. Moreover, the growing cells should predominate the culture and
thus shMiRs isolated from the growing cells should be over-represented.
234
Cells
number of T- number of
colonies
2
2
number
75cm
T-25cm
after 3
after
flasks
flasks
colonies after weeks in
DNA Pool selection
reseeded reseeded 3 weeks in T-75
T-25
3
3.75x10^5
7
none
0, 0, 1, 2, 3, 0, 1
n/a
4
2.44x10^5
4
none
1, 1, 0, 3
n/a
5
3.06x10^5
5
none
0, 2, 3, 0, 0
n/a
6
1.22x10^5
2
none
0, 0
n/a
7
8.50x10^5
8
none
0, 2, 2, 3, 1, 0, 1, 1
n/a
8
9.61x10^5
8
none
0, 1, 1,0 , 0, 0, 0, 0
n/a
9
9.61x10^5
8
none
2, 3, 3, 1, 0, 0, 1, 1
n/a
10
6.86x10^5
8
none
0, 0, 0, 0, 0, 0, 1, 1
n/a
11
12
p21
lamin
13
14
15
16
17
18
19
20
21
22
p21
lamin
23
24
25
5.11x10^5
3.75x10^5
1.64x10^5
3.39x10^5
2.22x10^5
2.28x10^5
1.08x10^5
1.22x10^5
3.61x10^4
6.10x10^4
1.03x10^5
1.56x10^5
9.72x10^4
1.94x10^4
1.36x10^5
4.11x10^5
2.47x10^5
4.08x10^5
3.42x10^5
8
7
1
4
4
4
2
2
none
1
2
3
1
none
2
4
4
7
6
none
none
none
none
none
none
none
none
2
none
none
none
2
1
none
none
1
1
none
0, 0, 2, 0, 0, 0, 1, 1
n/a
2, 1, 0, 4, 1, 0, 0
n/a
confluent
confluent
0, 0, 1, 1
n/a
6, 4, 8, 5
n/a
1, 0, 2, 0
n/a
2, 1
n/a
8, 11
n/a
n/a
0, 0
10
n/a
5, 6
n/a
0, 1, 0
n/a
5
0, 0
n/a
0
confluent
confluent
3, 1, 3, 4
n/a
3, 0, 1, 1
0
2, 0, 0, 0, 1, 1, 0
0
0, 0, 0, 1, 1, 0
n/a
26
27
28
29
30
31
32
p21
lamin
4.53x10^5
1.50x10^5
2.31x10^5
2.19x10^5
1.39x10^5
2.44x10^5
8.90x10^4
5.44x10^5
2.22x10^5
8
2
4
4
2
4
1
2
2
none
2
none
none
1
1
1
none
none
0, 0, 1, 3, 1, 4, 0, 0
n/a
4, 3
0, 1
1, 3, 1, 2
n/a
3, 0, 0, 1, 1
n/a
2, 4
1
1, 0, 0, 0
0
3
n/a
confluent
confluent
4, 1, 5, 6
0, 0, 0, 0
235
Cells
number of T- number of
number
75cm2
T-75cm2
after
flasks
flasks
DNA Pool selection
ressedeed
reseeded
33
1.40x10^4
none
1
34
2.31x10^5
4
1
35
2.25x10^5
4
none
colonies after
3 weeks in T-75
n/a
1, 0, 0, 0
0, 1, 0, 0
colonies
after 3
weeks in
T-25
0
0
n/a
36
37
38
39
40
41
42
p21
lamin
43
44
45
46
47
48
49
50
51
52
p21
lamin
53
4.81x10^5
1.89x10^5
2.19x10^5
3.06x10^5
1.22x10^5
1.83x10^5
3.25x10^5
2.05x10^5
4.30x10^5
4.72x10^4
7.20x10^4
7.20x10^4
7.50x10^4
8.05x10^4
1.65x10^4
1.35x10^4
5.80x10^4
4.70x10^4
4.72x10^4
1.67x10^5
4.25x10^5
3.17x10^5
9
3
4
5
2
3
6
2
4
none
1
1
1
1
none
none
none
none
none
2
4
6
none
1
none
none
none
1
none
none
4
2
1
1
1
1
1
1
3
2
2
none
4
none
54
55
5.61x10^5
3.00x10^5
9
5
none
none
2, 4, 1, 6, 1, 1, 0, 3
2, 2, 1, 1, 7
n/a
n/a
56
4.78x10^5
9
none
0, 5, 1, 2, 0, 6, 1, 2, 4
n/a
57
4.83x10^5
9
none
0, 0, 0, 0, 0, 0, 1, 0, 1
n/a
58
59
60
61
62
p21
lamin
4.42x10^5
3.74x10^5
4.08x10^5
3.67x10^5
1.94x10^5
2.00x10^5
6.92x10^5
8
7
7
7
4
2
4
none
none
none
none
none
none
4
0, 0, 0, 1, 2, 1, 4, 9
0, 2, 3, 1, 1, 6, 2
0, 4, 3, 5, 6, 1
3, 1, 4, 0, 0, 1, 0
1, 0, 3, 1
confluent
6, 2, 2, 2
n/a
n/a
n/a
n/a
n/a
n/a
n/a
236
0, 0, 0, 1, 0, 0, 0, 0, 0
n/a
0, 0, 1
0
0, 0, 0, 0
n/a
1, 0, 0, 0, 0
n/a
0, 0
n/a
2, 1, 1
0
0, 2, 0, 1, 0, 1
n/a
confluent
n/a
0, 0, 1, 0
0, 0, 0, 0
n/a
0, 0
0
0
2
0
0
2
3
0
n/a
0
n/a
0
n/a
0, 0, 0,
n/a
2, 0
n/a
1, 0
confluent
confluent
1, 1, 4, 6
0, 2, 0, 2
0, 1, 0, 0, 2, 7
n/a
DNA Pool
63
64
65
66
67
68
69
70
71
72
p21
lamin
73
number of
T-75cm2
flasks
Cells number
after selection ressedeed
8.60x10^4
1
6.90x10^4
1
5.50x10^4
1
8.88x10^4
1
1.33x10^5
2
2.50x10^5
4
1.00x10^5
2
6.10x10^4
1
9.40x10^4
1
1.03x10^5
2
8.75x10^5
4
1.13x10^6
4
2.86x10^5
5
number of
T-75cm2
flasks
reseeded
2
1
none
2
1
1
none
none
2
none
4
4
1
74
75
76
2.24x10^5
2.58x10^5
1.67x10^5
4
4
3
none
2
none
77
78
79
80
81
82
p21
lamin
83
84
85
86
87
88
89
90
91
92
p21
lamin
4.97x10^5
3.78x10^5
1.58x10^5
1.03x10^5
3.86x10^5
3.50x10^5
6.72x10^5
1.25x10^6
1.25x10^5
1.03x10^5
1.11x10^5
1.06x10^5
7.20x10^4
1.39x10^5
1.96x10^5
8.89x10^4
1.69x10^5
1.03x10^5
1.20x10^6
1.85x10^6
9
7
3
2
7
6
4
1
2
2
2
2
1
2
3
1
3
2
1
4
none
none
none
none
none
1
4
4
1
none
none
none
1
1
2
2
none
none
4
4
237
colonies
colonies after
after 3
3 weeks in T- weeks in T75
25
2
0, 0
1
1
0
n/a
2
3, 3
0, 2
0
0, 0, 1, 1
0
0, 0
n/a
0
n/a
2
0, 1
0,2
n/a
confluent
confluent
1, 6, 7, 2
2, 0, 0, 0
1, 0, 2, 2, 7
0
4, 2, 2, 3 :
n/a
pooled in 1
2, 3, 0, 4
0, 0
0, 1, 4
n/a
1, 3, 1, 3, 3, 3,
n/a
4, 2, 6
4, 3, 8, 3, 5, 6, 6
n/a
3, 5, 2
n/a
7,12
n/a
0, 0, 3, 2, 2, 0, 3
n/a
9, 3, 7, 3, 10, 11
4
confluent
confluent
1, 7, 1, 1
0, 1, 0, 1
0, 0
0
2,1
n/a
0, 0
n/a
0, 0
n/a
0
0
0, 1
0
0, 0, 1
0, 1
0
0, 0
0, 0, 0
n/a
0, 1
n/a
confluent
confluent
0, 0, 0, 2
0, 0, 0, 0
DNA Pool
93
94
95
96
97
98
99
100
1
2
p21
lamin
number of
Cells number T-75cm2
after
flasks
selection
ressedeed
3.78x10^5
7
5.84x10^5
7
3.11x10^5
5
2.11x10^5
4
2.19x10^5
4
3.98x10^5
2.99x10^5
2.47x10^5
1.08x10^5
2.16x10^5
1.18x10^6
9.75x10^5
7
5
4
2
4
1
4
number of
colonies
2
T-75cm
after 3
colonies after
3 weeks in T- weeks in Tflasks
75
reseeded
25
none
1, 1, 1, 0, 0, 0, 0
n/a
none
1, 1, 0, 0, 0, 0, 0
n/a
none
2, 0, 1, 1, 0
n/a
none
1, 4, 2, 2
n/a
none
1, 0, 1, 0
n/a
1, 0, 0, 0, 1, 1,
2
none
n/a
none
1, 0, 0, 0, 0
n/a
none
0, 0, 1, 0
n/a
none
0, 0
n/a
none
0, 0, 0, 0
n/a
none
confluent
confluent
none
1, 1, 0, 1
n/a
Red: flasks reseeded for genomic DNA extraction, if they are from the same pool they would have
been renamed pool X A, B, C, ect...
Table 5.3: Reseeding densities and number of growing colonies obtained after
growth
complementation assay
The table displays the pool number, the number of cells counted after puromycin selection, the number of
T-75 and T-25cm2 flasks reseeded and the number of growing colonies in each of them after 3 weeks at
38°C. The number in red indicate the flasks that were reseeded for genomic DNA extraction.
238
5.1.9 ShRNA constructs sequence recovery
Two distinct methods of target shRNA sequence retrieval were possible.
The first method uses a DNA barcoding system that was simultaneously developed by
both Berns and colleagues (Berns, Hijmans et al. 2004) and Paddison and colleagues
(Paddison, Silva et al. 2004). In the CSHL library, each shRNA construct was labelled
with a unique 60-nucleotide (nt) sequence such that each construct could be detected in a
process analogous to microarray analysis. This method was considered but would
require chips as well as setting up the labelling and scanning procedures. In addition, we
did not have much viral supernatant for extra optimisation, so this method was not
selected.
The second method involved extracting genomic DNA from the growing CL3 EcoR cells
followed by PCR amplification using vector-specific primers that spanned the shRNA
insert sequence and cloning into a TA-cloning vector. Sequencing of multiple colonies
would allow determination of the identity of the functional shRNA species integrated
within the cells.
200 ng of genomic DNA was used for PCR amplification with the pSM2 specific
primers: pSM2 longF and pSM2 longR, and the separated by electrophoresis on a 3%
agarose gel with ethidium bromide alongside a positive control for PCR, namely 5 l of
the PCR product generated from the amplification of 100 ng of pSM2 scrambled control
vector. The DNA was generally not visible after this first round of PCR so a second
round of amplification using a set of nested primers, namely pSM2 F and pSM2 R, that
were internal to the first set of primers was used to amplify the inserts before TOPOcloning. This time analysis by electrophoresis of 5μl of PCR product revealed a product
of 424 bp in all samples that corresponded to the expected insert sequence, but not in a
negative control sample where water was substituted for template DNA. The PCR
product was cloned into pCR2.1-TOPO vector (Invitrogen) using the TOPO TA Cloning
239
Kit (Invitrogen). The resulting transformed E. coli were plated onto LB-agar plates
containing 50 g/ml final concentration ampicillin and 80 l of 20 mg/ml X-gal and
incubated at 37C overnight. Blue/white selection was used to identify positive clones
from which, plasmid DNA was subsequently extracted using the QIAprep Spin
Miniprep kit (QIAGEN). This step insured that every bacterial clone picked would only
contain one insert. Plasmid DNAs were sequenced by the MRC Prion Unit sequencing
facility.
The sequencing was performed with M13R primer which is a primer specific for the
pCR2.1-TOPO vector. Each DNA was sequenced to reveal the shRNAmir insert.
The insert was determined by searching for the miR-30 context and miR-30 loop (Figure
5.1) that are common to all inserts and frame the hairpins. The sequence of the hairpin
was used to identify the gene by searching the pSM2 database provided by Open
Biosystems but also by BLASTN analysis of the NCBI human genome database. The
sequences that could not be linked back to the list of insert sequences in the Open
Biosystems were not pursued and are not presented here.
5.1.10 Results of the primary screen
The rescued shRNAmirs hairpins identified 111 different genes and another 30 inserts
corresponding to unidentified loci. For each pool, the number of time sequences were
obtained, the corresponding insert references and gene symbols are shown in Table 5.4.
This table also shows in the last column if the insert recovered was a match to the inserts
present in that particular pool (indicated by ―match‖) or an insert listed from another
pool of the library (indicated by ―listed in pool X‖). 24 different inserts did not come up
in the pool they were supposed to. Some inserts were detected several times in multiple
other pools. For example, the insert V2HS_119967 which was listed as an insert in pool
52 came up several times in pools 3a, 4, 7a, 9a and b, 12a and b, 30a and b, 59b, 60b and
c, 64a, 71, 72, 74, 77, and 98d.
240
Pool
insert reference
gene names
3a
3a
V2HS_63142
V2HS_119967
3a
V2HS_56766
3a
V2HS_95607
leucine rich repeat containing 37A
3b
3b
4
V2HS_53974
V2LHS_63482
V2HS_119967
PRO0255 protein
paired box 1
LOC100287210
4
V2HS_66751
4
4
V2HS_70011
V2HS_98079
serum amyloid A-like 1
human solute carrier family 22
5a
V2HS_108647
human LOC349868
5a
V2HS_70473
5a
V2HS_71958
5b
V2HS_119967
5b
V2HS_66652
keratin associated protein 5-9
LOC100287210
acyl-CoA synthetase medium-chain
family member 3
polymerase (dna directed), mu
POLM
2
listed in pool 79
listed in pool
330
human olfactory receptor, family 5,
subfamily P, member 3
OR5P3
3
match
7
match
human protein phosphatase 3 (formerly
2B), catalytic
V2HS_98079
7a
V2LHS_97017
V2HS_119967
9a
V2HS_62506
9b
9b
9b
9b
V2HS_55950
V2HS_119967
V2HS_70312
v2HS_71740
9b
V2HS_98079
9c
V2HS_64384
5
2
2
3
7a
V2HS_108647
PRO0255
PAX1
LOC100287210
listed in pool
402
match
match
listed in pool 52
LOC349868
human cyclin-dependent kinase 8
LOC100287210
CD28 antigen
solute carrier family 22 (extraneuronal
monoamine transporter), member 3
9a
1
match
listed in pool 79
polymerase (dna directed), mu
9a
LRRC37A
match
V2HS_70473
v2HS_69776
match
2
1
V2HS_112910
V2HS_119967
V2HS_64878
7b
1
4
7a
7a
7a
v2HS_56367
ACSM3
SAAL1
SLC22
5b
7b
number of
sequences
KRTAP5-9
3
match
LOC100287210
4
listed in pool 52
gene symbol
sterile alpha motif containing 4a
PPP3CB
1
POLM
4
CDK8
LOC100287210
CD28
1
1
1
SLC22A3
1
SAMD4
1
human similar to progesterone receptor membrane
component
cytochrome P450, family 4, subfamily Z,
CYP4Z2P
polypeptide 2
6
1
match
match
LOC100287210
1
listed in pool
79
listed in pool 52
FAM181B
1
match
PSCA
LOC100287210
TFG
ATXN10
2
3
4
1
match
listed in pool 52
match
SLC22A3
1
APBA2
1
1
LOC100287210
family with sequence similarity 181,
member B
prostate stem cell antigen
LOC100287210
TRK-fused gene
ataxin 10
solute carrier family 22 (extraneuronal
monoamine transporter), member 3
amyloid beta (A4) precursor protein
binding family A member 2
match
listed in pool
330
match
listed in pool 52
listed in pool 16
listed in pool
79
listed in pool
172
listed in pool
79
match
SAMD4
1
LOC100287210
SIRPB2
5
3
PTPN13
1
match
V2HS_101484
LOC100287210
human signal-regulatory protein beta 2
protein tyrosine phosphatase, nonreceptor type 13
doublecortin domain containing 2B
listed in pool
172
listed in pool 52
match
DCDC2B
4
listed in pool 74
V2HS_119967
LOC100287210
LOC100287210
5
13a
V2HS_162164
iodotyrosine deiodinase
IYD
1
listed in pool 52
listed in pool
839
13a
V2HS_55731
phenylalanine-tRNAsynthetase-like,
beta subunit
FARSB
1
9c
V2LHS_97017
sterile alpha motif containing 4a
12a
12a
V2HS_119967
V2HS_58950
12b
V2HS_57276
12b
12b
241
match
Pool
insert
reference
13a
13a
V2HS_59258
V2HS_59891
13a
V2HS_71174
13b
V2HS_57692
13b
13c
13d
V2HS_68714
V2HS_59653
V2HS_55731
13d
V2HS_64320
13d
v2HS_71174
13d
16a
18
V2HS_71453
V2HS_64878
V2HS_59716
18
19a
V2HS_57051
VH2S_106158
V2HS_59560
v2HS_55310,
v2HS_55312
V2HS_68437
V2HS_119967
30a
30a
30b
30b
30b
30b
32
41
54a
54a
54a
54b
54c
54c
55a
55a
55b
55b
56a
56b
dynein, light chain, roadblock-type 1
TAO kinase 1
peroxisome proliferator-activated receptor
gamma,coactivator 1 alpha
human similar to peptidase (prosome,
macropain) 26S subunit
abl-interactor 1
phenylalanyl-tRNA synthetase, beta subunit
Smith-Magenis syndrome chromosome region,
candidate 7
peroxisome proliferator-activated receptor
gamma
CD28 antigen
StAR-related lipid transfer (START) domain
containing 6
V2HS_247318 TMEM9 domain family, member B
19a
19b
21
30a
gene symbol
v2HS_63989
19a
21
gene names
cDNA DKFZp564H0764
chromosome 13 open reading frame 15
glucosamine-phosphate N-acetyltransferase 1
DYNLRB1
Taok1
match
match
PPARGC1A
1
match
LOC643766
2
match
ABI1
FARSB
3
1
1
match
match
match
SMCR7
1
match
PPARG
3
match
CD28
2
8
1
match
match
match
STARD6
1
match
TMEM9B
DKFZp564H076
4
4
match
5
match
c13orf15
1
10
listed in pool 82
match
GNPNAT1
6
match
2
6
match
listed in pool 52
7
match
1
1
1
listed in pool 149
listed in pool 52
match
8
match
1
match
5
6
1
2
1
6
2
1
match
match
match
match
match
match
match
match
2
match
1
1
6
1
4
match
listed in pool 58
cDNA FLJ30947
FLJ30947
LOC100287210
LOC100287210
human protein phosphatase 4, regulatory
V2HS_34338
PPP4R2
subunit 2
V2HS_114455 testis derived transcript (3 LIM domains)
TES
V2HS_119967 LOC100287210
LOC100287210
V2HS_34338 protein phosphatase 4, regulatory subunit 2
PPP4R2
solute carrier family 33 (acetyl-CoA transporter),
V2HS_36467
Slc33a1
member 1
ubiquitin-activating enzyme E1C (UBA3
V2HS_46793
UBA3
homolog, yeast)
v2HS_42104 cDNA FLJ38187
FLJ38187
v2HS_48278 choline kinase-beta
CHKB
V2HS_112629 basic transcription factor 3, like 1
BTF3L1
V2HS_121153
V2HS_125538
V2HS_129527
V2HS_112629 basic transcription factor 3, like 1
BTF3L1
V2HS_129417
SH3 domain binding glutamic acid-rich protein
v2HS_117465
SH3BGRL2
like 2
V2HS_119051
V2HS_116174 YTH domain containing 2
YTHDC2
V2HS_119120 hypothetical protein FLJ20032
FLJ20032
V2HS_117914 similar to transketolase (DKFZP434L1717)
TKTL2
v2HS_115231 rab23 member RAS oncogene family
RAB23
242
number
of
sequence
s
3
5
match
match
Pool
insert
reference
gene names
gene symbol
56b
56c
56c
56c
58a
58a
V2HS_117914
V2HS_117239
V2HS_120429
V2HS_94458
V2HS_116174
v2HS_118722
58b
v2HS_112838
58b
V2HS_112982
58b
V2HS_116174
58c
v2HS_112838
58c
58c
58c
59a
66c
71
71
71
V2HS_118254
V2HS_122548
V2HS_125075
v2HS_111554 interleukine 2
small glutamine-rich tetratricopeptide repeat
V2HS_176550
(TPR)-containing, beta
V2HS_116377 TMEM135 domain family
v2HS_117064 CD1e molecule
V2HS_119967 LOC100287210
olfactory receptor, family 8, subfamily K,
V2HS_120757
member 1
V2HS_108647
v2HS_115544 DEAD (Asp-Glu-Ala-Asp) box polypeptide 47
intermediate filament protein syncoilin
V2HS_116833
(SYNCOILIN),
V2HS_121585 dual specificity phosphatase 3
V2HS_117903 glutamine rich 2
V2HS_119967 LOC100287210
V2HS_121585 dual specificity phosphatase 3
V2HS_128131
solute carrier family 22 (extraneuronal
V2HS_98079
monoamine transporter), member 3
V2HS_119967 LOC100287210
V2HS_121585 dual specificity phosphatase 3
V2HS_119967
V2HS_121013 plasma kallikrein-like protein 4
latent transforming growth factor beta binding
V2HS_117673
protein 3
V2HS_115659 three prime repair exonuclease 1
latent transforming growth factor beta binding
V2HS_117673
protein 3
V2HS_115659 three prime repair exonuclease 1
V2HS_103818 LOC284804
V2HS_119967 LOC100287210
V2HS_97891 melanoma antigen (LOC51152)
72
V2HS_106385 MGC30618
59a
59b
59b
59b
59b
59c
59c
59c
60a
60b
60b
60b
60b
60b
60c
60c
64a
64b
66a
66b
66b
similar to transketolase (DKFZP434L1717)
human chromosome 9 open reading frame 58
TKTL2
C9orf58
arachidonate 15-lipoxygenase, type B
human FLJ21940 protein
layilin
ectonucleoside triphosphate
diphosphohydrolase 3
human chromodomain helicase DNA binding
protein 3
human FLJ21940 protein
ectonucleoside triphosphate
diphosphohydrolase 3
WD repeat and FYVE domain containing 2
ribosomal protein S3A
ALOX15B
FLJ21940
LAYN
243
number
of
sequenc
es
2
match
3
match
1
match
2
listed in pool 146
1
match
2
match
ENTPD3
3
match
CHD3
1
match
FLJ21940
1
match
ENTPD3
1
match
WDFY2
RPS3A
IL2
1
4
1
8
match
listed in pool 53
match
match
SGTB
2
match
TMEM135
CD1E
LOC100287210
1
3
4
match
match
listed in pool 52
OR8K1
1
match
DDX47
1
3
listed in pool 79
match
SYNC
1
match
DUSP3
QRICH2
LOC100287210
DUSP3
10
4
1
2
2
match
match
listed in pool 52
match
match
SLC22A3
1
listed in pool 79
LOC100287210
DUSP3
Klkbl4
8
2
10
3
listed in pool 52
match
listed in pool 52
match
LTBP3
15
match
TREX1
5
match
LTBP3
8
match
TREX1
LOC284804
LOC100287210
LOC51152
1
2
9
5
match
listed in pool 78
listed in pool 52
match
MGC30618
1
match
Pool
insert
reference
gene names
gene symbol
number of
sequences
72
V2HS_119967 LOC100287210
LOC100287210
9
listed in pool 52
74
V2HS_119967 LOC100287210
LOC100287210
11
listed in pool 52
77
78a
78a
78b
V2HS_119967
V2HS_94763
V2HS_99138
V2HS_102207
LOC100287210
PRRX1
SLC25A21
TMEM63B
12
1
7
1
listed in pool 52
match
match
match
78b
78b
78b
78c
V2HS_103818 LOC284804
v2HS_95356 armadillo repeat containing, X-linked 2
v2HS_98650 mitochondrial ribosomal protein 63
V2HS_102155
LOC284804
ARMCX2
MRP63
4
2
1
1
match
match
match
match
78c
78c
V2HS_108506
V2HS_96236
ZBTB1
1
2
78d
V2HS_184999 eukaryotic initiation factor 4A isoform 1
EIF4A3
3
78d
79a
V2HS_96236
V2HS_105974
ZBTB1
1
3
79a
V2HS_184999 eukaryotic initiation factor 4A isoform 1
EIF4A3
3
match
match
listed in pool
528
match
match
listed in pool
528
SLC22A3
4
match
2
match
SLC22A3
5
match
ANTXRL
1
match
SLC35F4
2
match
79a
V2HS_98079
79b
V2HS_108647
LOC100287210
paired related homeobox 1
solute carrier family 25
TMEM63 domain family member B
zinc finger and BTB domain containing 1
zinc finger and BTB domain containing 1
solute carrier family 22 (extraneuronal
monoamine transporter), member 3
79c
solute carrier family 22 (extraneuronal
V2HS_98079
monoamine transporter), member 3
V2HS_101943 anthrax toxin receptor-like
79c
V2HS_105093 human solute carrier family 35, member F4
79c
V2HS_106291
1
match
79c
V2HS_106472
1
match
79c
79c
NCOR1
1
4
match
match
OR5D3P
1
match
80a
V2HS_107395
v2HS_91777 nuclear receptor co-repressor 1
olfactory receptor, family 5, subfamily D,
V2HS_91794
member 3 pseudogen
v2HS_102441 adenylate cyclase 1
ADCY1
5
match
80a
80a
80b
V2HS_130882 glutamate receptor, metabotropic 3
v2hs_97368
yippee-like 5
v2HS_102441 adenylate cyclase 1
GRM3
YPEL5
ADCY1
1
2
5
80b
V2HS_106345
79b
79c
v2HS_99202
82a
V2HS_184999 eukaryotic initiation factor 4A isoform 1
82a
V2HS_101845 human prickle-like 2 (Drosophila)
82a
V2HS_109096
V2HS_99423
PRICKLE2
3
1
listed in pool 84
1
match
1
match
3
match
1
5
match
match
similar to heterogeneous nuclear ribonucleoprotein A1
hypothetical protein DKFZp434K1172
82b
V2HS_10959
6
V2HS_108399
82b
82b
V2HS_146196 shisa homolog 7
v2HS_93536 human proteolipid protein 1
82b
1
4
80b
82a
EIF4A3
listed in pool 91
match
match
listed in pool
144
match
listed in pool
528
match
DKFZp434K117
2
shisa7
PLP1
244
1
Pool
insert
reference
gene names
gene symbol
82b V2LHS_97017 sterile alpha motif domain containing 4A
3
listed in pool
172
match
TP53
1
match
RASA4
1
match
BCL2L12
1
SAMD4
82c V2HS_106158
82c
v2HS_93615
p53
82c
v2HS_95112
RAS p21 protein activator 4
82c
V2HS_99526
BCL2 like 12
82c V2LHS_97017 sterile alpha motif domain containing 4A
82d
V2HS_94640
aryl hydrocarbon receptor nuclear
translocator-like
82d V2HS_100174
desmoglein 4
82d
V2HS_96026
adnp homeobox2
82d
vhs_100819
Rho GTPase activating protein 20
v2HS_95019
84a
V2HS_97152
84b
v2HS_95019
zinc finger protein 16
biquitin-conjugating enzyme E2, J1 (UBC6
homolog
zinc finger protein 16
1
match
SAMD4
4
ARNTL
3
DSG4
2
ADNP2
3
listed in pool
172
listed in pool
411
listed in pool
145
match
ARHGAP20
2
match
1
match
ZNF16
9
match
UBE2J1
1
match
ZNF16
10
match
84a V2HS_106409
84a
Number
sequence
94
V2HS_141495
zinc finger protein 454
ZNF454
7
listed in pool
96
94
V2HS_33370
leucine zipper protein 1
LUZP1
1
match
RSPH10B
6
match
KCNJ2
1
95a V2HS_130457
95a V2HS_131154
radial spoke head 10 homolog B
(Chlamydomonas)
KCNJ2, potassium inwardly-rectifying
channel, subfamily J
95c V2HS_184999
eukaryotic initiation factor 4A isoform 1
98a V2HS_141367
cDNA FLJ37626
98d V2HS_100218
human deltex 3 homolog (Drosophila)
98d V2HS_119967
LOC100287210
98d V2HS_133299
98d V2HS_135564
human insulin-like growth factor binding
protein 6
Homo sapiens chromosome X open reading
frame 57
98d V2HS_146491
match
listed in pool
528
match
EIF4A3
1
FLJ37626
10
DTX3
4
LOC100287210
2
IGFBP6
3
match
CXORF57
1
match
1
match
listed in pool
78
listed in pool
52
Table 5.4: Results of the screen
The table displays the genes that were recovered by sequencing the genomic DNA extracted from the
growing colonies. The columns represent the pool that gave rescue, the name and symbol of the genes
recovred by sequencing from that pool, the number of time this sequence was recovered and their
expected location in the library.
245
This suggested that there may have been cross-contamination of the library perhaps
during replica plating. However, this was not a problem since the aim of this screen was
to identify any shRNAmiRs that can bypass the growth arrest. Since this insert originally
from pool 52 was isolated from so many pools, it could be that it was a strong positive or
that it was highly represented within the library.
Pools 13, 78 and 82 that produced more colonies and colonies that were larger and
healthier than others, identified the following genes:
Pool 13: IYD, DYNLRB, FARSB, PPARGC1A, Taok1, ABI1, LOC643766, FARSB,
PPARG, SMCR7;
Pool 78: PRRX1, SLC25A21, ARMCX2, LOC284804, MRP63, TMEM63B, ZBTB1,
EIF4A3, ZBTB1;
Pool 82: DKFZp434K1172, PRICKLE2, SAMD4, PLP1, shisa7, SAMD4, BCL2L12,
RASA4, TP53, ARNTL, DSG4, ADNP2, ARHGAP20.
It is interesting to note that pool 82, one of the pools that gave the best rescue of the
screen, contained the shmiRs targeting TP53, one of which (V2HS_93615) was
identified in the screen, thereby validating it. shRNAmirs targeting p21 were not present
in this library.
Unfortunately the other shMIRs targeting TP53 were not isolated suggesting that either
my screen was not saturating or that the other p53 shRNAmirs were unable to silence
p53 at a level sufficient to bypass senescence in CL3EcoR cells.
246
5.2 IN VITRO VALIDATION OF THE SCREENING
5.2.1 Overlap of the candidates of the shRNA screen with microarray data for
genes up-regulated upon senescence in CL3EcoR cells
To prioritise the 137 candidates identified from the primary screen for functional
validation, they were compared to genes found to be up-regulated upon senescence and
whose expression was reversed when senescence was abrogated (ER et al, submitted).
This identified 5 common genes, ATXN10, LAYN, LTBP3, SGTB and TMEM9B. The
microarray expression profiling data is presented in Table 2. They were all up-regulated
upon senescence growth arrest: ATXN10 by 1.3 fold (p-value 8.3x10-4), LAYN by 2
fold (p-value 1.8x10-4), LTBP 3 by 1.32 (p-value 8.6x10-8) and 0.45 (p-value 6.1x10-5),
SGTB by 1.3 fold (p-value 8.8x10-4) and TMEM9B by 1.4 (p-value 1.4x10-9 ) and 1.3
fold (p-value 2.2x10-4) respectively.
The identification of TMEM9B was particularly remarkable because the microarray
analysis had suggested that senescence growth arrest in these cells was associated with
activation of the NF-κB signalling pathway and TMEM9B had previously been shown
to be able to activate NF-κB dependent reporter constructs (Matsuda, Suzuki et al. 2003;
Dodeller, Gottar et al. 2008). TMEM9B is also essential for activation of NF-κB by
TNF and acts downstream of RIP1 and upstream of MAPK and IκB kinases at the level
of TAK1 (Dodeller, Gottar et al. 2008).
5.2.2 Optimisation of the GIPZ lentiviral library
The shmiRs from the pGIPZ and the pSM2 library are essentially similar, what is
different is the vector system and how they are delivered to the cells. Moreover, it was
observed that the pSM2 vectors were not stable in E.coli and that the lentiviral shmiRs
247
were much more stable. We chose to use lentiviral vectors since they were freely
available to us through the RNAi consortium rather than the pSM2 library.
For the lentiviral supernatants to be used for the secondary screen, optimisation of the
packaging and infection protocols was necessary. The packaging was done in HEK293
cells as suggested by protocols provided by the RNAi consortium. The infection itself,
however, required optimisation as getting a level of silencing high enough to see a
biological phentotype proved to be tricky.
Katharina Wanek had been working with the lentiviral shmiRNA library and found that
selection at 2µg/ml puromycin for 4 days, failed to give a satisfactory expression of the
vector, monitored by the level of GFP positive cells. In addition, even shmiRNAs
targeting p21could not yield rescue. What was more important is the pRS-p21F shRNA
selected at 2µg/ml gave excellent rescue, therefore the problem laid with the shmiR
vector.
However, she found that a higher level of puromycin, 6µg/ml enriched for infected cells
in which the GIPZ vector was expressed at higher levels. Furthermore, maintaining
selection upon shift to 38°C gave a lower level of background reversion. Others have
found that enriching for the higher expression by GFP expression also yields higher
levels of RNA knockdown.
5.2.3 Secondary screen using lentiviral shRNA silencing
In order to validate these 5 targets, the secondary screen was performed in two parts:
First, complementation was attempted with a mix of multiple (as many as were
available) silencing constructs from the Human GIPZ lentiviral shRNAmir library
available for these genes. This was then repeated using each construct individually.
To determine if silencing of TMEM9B would bypass senescence, 4 GIPZ lentiviral
silencing
constructs
(V2LHS_247318,
248
V2LHS_58957,
V2LHS_58958
and
V2LHS_58959) targeting TMEM9B were obtained, pooled and introduced into CL3EcoR
cells after packaging as lentiviruses. Lentiviral human GIPZ LaminA/C shRNAmir
(V2LHS_62719) was used as a negative control. The transduced cells were selected
with 6µg/ul of puromycin for 5 days and reseeded in triplicate at 0.5x105 in T-75 flasks.
The cells were then shifted to 38°C for 3 weeks, stained with methylene blue and the
number of colonies was counted. Silencing of TMEM9B was clearly able to overcome
senescence (Figure 5.6). The result shows small background with the Lamin A/C
constructs with an average of approximately 20 small colonies and a much larger
number of healthy growing colonies with the mixed silencing constructs with an average
of approximately 100 colonies. This confirmed the results for TMEM9B presented in
the Chapter 4 and showed that it was possible to achieve silencing with the GIPZ
lentiviral vector in the CL3 EcoR cells.
Moreover, each of the constructs was able
overcome senescence arrest when introduced individually into CL3 EcoR cells, with
V2LHS_58957 being the most efficient (Figure 5.7); it was interesting to note that the
least efficient hairpin (247318), was the hair pin isolated by the shRNA screen.
Next, only individual shRNA inserts were used for the genes LTBP3, ATXN10, LAYN
and SGTB.
The complementation assay for LTBP3 silencing show the results for only one silencing
construct (V2HS_34089) that was available at the time. The LTBP3 silencing construct
(Figure 5.8) produced numerous healthy growing colonies in comparison to the Lamin
shRNAmiRs which resulted in low level background.
ATX10 silencing was tested with 4 different silencing constructs namely V2HS_71735,
V2HS_71736, V2HS_71737 and V2HS_71740. All four constructs gave a high level of
rescue when compared to the negative control Lamin silencing construct (Figure 5.9).
249
Figure 5.6:Silencing of TMEM9B (mix)
CL3EcoR cells were infected in triplicate with a mix of lentiviruses shRNAmir silencing constructs for
TMEM9B (human GIPZ lentiviral shMIR V2LHS_247318, V2LHS_5895, V2LHS_58958 and
V2LHS_58957) and assayed for growth complementation at 38°C. After 3 weeks the number of growing
colonies were counted.
250
Figure 5.7: Silencing of TMEM9B (individual)
CL3EcoR cells were infected in triplicate with lentiviruses expressing the indicated shRNAmir silencing
constructs and assayed for growth complementation at 38°C. After 3 weeks the numbers of growing
colonies were counted.
251
Figure 5.8: Silencing of LTBP3
CL3EcoR cells were infected in triplicate with lentiviruses expressing the indicated shRNAmir silencing
constructs and assayed for growth complementation at 38°C. After 3 weeks the numbers of growing
colonies were counted.
252
Figure 5.9:Silencing of ATXN10
CL3EcoR cells were infected in triplicate with lentiviruses expressing the indicated shRNAmir silencing
constructs and assayed for growth complementation at 38°C. After 3 weeks the number of growing
colonies were counted.
253
SGTB silencing was tested with 3 different silencing constructs namely V2HS_176551,
V2HS_218863, and V2HS_176555. Two out of three constructs gave a high level of
rescue when compared to the negative control Lamin silencing construct (Figure 5.10).
The last construct did not permitted bypass of the growth arrest significantly.
LAYN silencing was tested with 2 different silencing constructs namely V2HS_265009,
and V2HS_118722. Both constructs yielded numerous growing healthy colonies when
compared to the negative control Lamin silencing construct (Figure 5.10).
Taken together the results showed that silencing of TMEM9B, ATXN10, LAYN,
LTBP3 and SGTB were able to bypass senescence in the conditionally immortal human
fibroblasts.
254
Figure 5.10: Silencing of SGTB and LAYN
CL3EcoR cells were infected in triplicate with lentiviruses expressing the indicated shRNAmir silencing
constructs and assayed for growth complementation at 38°C. After 3 weeks the numbers of growing
colonies were counted.
255
5.3 DISCUSSION
To directly identify the downstream effectors of the p53-p21 and p16-pRB pathways
crucial for mediating entry into senescence, I have carried out a loss-of-function RNA
interference screen in the conditionally immortal HMF3A human fibroblasts.
These
cells are immortal but undergo a rapid irreversible senescence arrest which can be
readily bypassed upon inactivation of the p53-p21 and p16-pRB pathways. This screen
identified 111 known genes and another 30 shRNAmirs corresponding to unidentified
loci. Comparison of these known targets with genes up-regulated upon senescence in
these cells identified 5 common genes TMEM9B, ATXN10, LAYN, LTBP3 and SGTB.
Direct silencing of these 5 genes using lentiviral shRNAmirs bypasses senescence in the
HMF3A cells. Although none of these five genes had previously been linked to cellular
senescence, TMEM9B has been suggested to be an upstream positive modulator of NFκB and I have found that activation of NF-κB signalling acts to promote senescence.
5.3.1
Sensitivity, Stringency and Saturation
The effectiveness of any screen is dependent upon its sensitivity; therefore, it was
important to minimise the background levels of false-positive hits without losing
information concerning the identity of all true positive hits. In this respect, optimal
conditions for performing an RNAi screen in the HMF3A system had been previously
determined by the development of the CL3 EcoR complementation assay. The sensitivity
of the screen itself was then tested successfully using complementation assay with a
spiked mixture of positive shRNA construct namely pRS p21 shRNA into negative
control namely pRS Lamin A/C, at a ratio of 1 in 200 (Figure 5.5). This demonstrated
that the screen assay was sensitive enough to be able to discriminate the effects of a
single shRNA constructs in a mix of 200. The stringency of the screen also proved
satisfactory, with a low level of reversion (Figure 5.5, A and D).
256
siRNAs can have 'off-target' effects, which are often the result of partial homology to
other transcripts (Jackson, Bartz et al. 2003; Semizarov, Frost et al. 2003). The Open
Biosystems shRNA library was designed to avoid off-target effects by minimizing
homology of shRNAs to other transcripts and by offering usually more than one
constructs per gene to silence. It is very unlikely that two independent siRNAs against
the same transcript target a common off-target transcript for suppression.
CL3EcoR complementation assay with p53 silencing was an easy way to validate the
screen with a positive control. A shRNA for TP53 was present in pool 82. Infection of
the cells with the viral supernatant containing constructs of the pool 82 gave a successful
rescue from growth arrest compared to the negative control and the p53 shRNA hairpin
was identified by sequencing, although only 1 insert for p53 was recovered. This result
does reinforce the quality of the assay. I did not identify the shRNA for
p21CIP1/WAF1/Sdi1 in the genetic screen as it was not present in the RNAi library. The fact
that only one shMiRs targeting p53 was isolated indicated that the first screen might not
be saturating.
The screen is unlikely to be saturating since all the shRNA targeting the same gene were
not always isolated from a positive pool. For example, for all the silencing constructs
tested here, namely TMEM9B, SGTB, LAYN and ATX10 shRNAs, all shown to rescue
in the CL3EcoR cells, only one hairpin each was recovered from sequencing out of 4, 3, 2
and 4 inserts respectively represented in their respective pools. This in accordance with
that of Westbrook and colleagues (Westbrook, Stegmeier et al. 2005), who similarly
raised concerns over the lack of saturation in shRNA screens.
These characteristics have to be taken into consideration when examining the results of
the screen. The CL3 EcoR system was suitable for the application of such an in vitro RNAi
screen with the final aim of identifying novel genes that are downstream effector of the
p53-p21 and p16-pRb pathways. Indeed, these cells were highly infectable, yielded low
background and grew very well.
257
In addition, it is important to note here that these cells were thawed for each assay of 10
pools (infected on the same week) from a unique batch of frozen CL3EcoR. This
minimised the background by limiting any potential reversion due to a long term in vitro
culture. In fact, it was found in a first trial of the screen that when passaged extensively
in stressing conditions (LT inactivated), the CL3 EcoR cultures can acquire mutant cells
than can become enriched upon cultivation and affect the outcome of the screen by
increasing the background.
5.3.2 Positive hits of the primary HMF3A retroviral shRNA screen
The primary screen shows that a total of 34 pools out of 100 gave growing colonies at a
level above background. Particularly, the pools 13, 78 and 82 gave a higher number of
colonies and colonies of a larger size. For each pool considered as a hit, 1 to 4 flasks
containing the highest number of colonies were reseeded for genomic DNA extraction
resulting in a total of 81 sub-pools to analyse for proviral shRNAmiRs. Sequencing
revealed a match for 111 different genes and another 30 unidentified loci.
A number of sequences were never matched to either the OpenBiosystems hairpin
database or the NCBI genomic database. This could be due to the removal of the hairpin
from the bank between the release of the pools and the sequence analysis or to a genetic
mutation making the sequence impossible to recognize. Furthermore, short sequences
are difficult to match against the all human genome and only one base mutation could
render the sequence unrecognizable. The unknown sequences could also correspond to
expression tags not yet documented as the library contain both known ESTs and
unidentified expression tags.
Due to time limitation, it wasn‘t possible to run a complete secondary screen for the
gene list since for each gene there were at least another 1 to 3 shMiRs available in the
pGIPZ library. For this reason, a filtering was performed by overlapping the microarray
up-regulated genes with the primary screen gene list.
258
5.3.3 Overlap with the microarray up-regulated genes reveals new targets
The list of candidates identified from the primary screen was overlapped with the
microarray up-regulated genes whose expression was reversed upon complementation.
The overlap was 5 genes: ATX10, LAYN, LTBP3, SGTB and TMEM9B. The list could
be longer if the all microarray data was compared, however, for more consistency, only
the differential data set containing ~8000 significantly differentially expressed genes as
described in Chapter 4 was compared to the candidates. These five genes were then
silenced in a secondary screen using lentiviral shRNAmirs and were able to bypass
growth arrest.
5.3.4 In vitro validation of ATXN10, LAYN, LTBP3, SGTB and TMEM9B
silencing
5.3.4.1 TMEM9B
Pool 19 of the primary shRNA screen identified TMEM9B as a target which upon
silencing would result in bypass of senescence arrest. TMEM9B was up-regulated 0.44
and 0.39 log2 fold change (P value 1.47x10 -7 and 1.97x10-4, Table 5.5) upon growth
arrest in the CL3EcoR cells which was reversed upon abrogation of the p53-p21 or the
p16-pRB pathways. TMEM9B expression was unaffected by quiescence. A role for
TMEM9B in inducing senescence was further supported by my finding that direct
silencing of TMEM9B using 4 different lentiviral shRNAmirs either as a mix or
individually bypasses senescence in the HMF3A cells (Figure 5.6 and 5.7).
259
Probe
Symbol
GA
Q
HS wt_LT
GSE
p53
pRS p53 pRS_p21 E1A
E7
E2F
DB
208832_at ATXN10 0.36
-0.64
0.25
0.11
0.34
-0.18
-0.35
0.11
0.02
-0.23
228080_at
LAYN
1.04
-0.56
0.30 -0.84
-0.56
-0.95
-0.70
-1.56
-0.56
-0.11
219922_s_at LTBP3
1.32
0.44
0.66 -0.43
-0.31
-1.04
-1.32
-1.02
-1.19
-1.41
227308_x_at LTBP3
0.46
-0.14
0.03 -0.19
-0.12
-0.43
-0.57
-0.44
-0.28
-0.55
0.36
-1.02
0.29 -0.85
0.03
-0.42
-0.81
-0.75
-0.56
-0.69
218065_s_atTMEM9B 0.44
228745_at
SGTB
-0.27 -0.11 -0.43
-0.29
-0.33
-0.38
-0.13
-0.20
-0.18
222507_s_atTMEM9B 0.39
-0.29 -0.19 -0.35
-0.29
-0.28
-0.34
0.11
-0.12
-0.07
Table 5.5: Senescence specific changes with complementation for ATXN10, LAYN, LTBP3, SGTB
and TMEM9B
Log2 fold changes in gene expression (and their p-values) that occur upon irreversible growth arrest (GA),
heat shock(HS), quiescence(Q) and the indicated complementation. Up-regulated transcripts are indicated
in green whereas down-regulated transcripts are in red. Results for ATXN10, LAYN, LTBP3, SGTB and
TMEM9B are shown .
260
TMEM9B is a glycosylated protein localized in membranes of the lysosome and
partially in early endosomes. TMEM9B has also been shown to be an important
component of TNF signalling and a module shared between the interleukin-1 and Tolllike receptor pathways. It was also shown to be essential for the TNF activation of both
NF-κB and MAPK signalling pathways by acting downstream of RIP1 and upstream of
the MAPK and IκB kinases at the level of the TAK1 complex (Dodeller, Gottar et al.
2008). It has also been identified by a large-scale characterization study to be one of the
genes activating NF-κB and MAPK signalling pathways (Matsuda, Suzuki et al. 2003).
These results are all consistent with my finding that in the conditionally immortal
HMF3A cells, senescence growth arrest is associated with an activation of NF-B
signalling and suppression of this pathway bypasses senescence (Chapter 4). These
results also seem to suggest that TMEM9B up-regulation could be the cause of the NFB activation upon senescence.
The details of this activation are not known, nor are the ways in which the NF-κB
pathway is involved in triggering senescence here. Investigating further the details of
this involvement could include some expression analysis of the cells expressing the NFκB signalling or not in order to determine which genes are affected and particularly
whether the p16-pRb and p53-p21 pathways are affected by it.
5.3.4.2 LTBP3
The latent TGF--binding protein 3 (LTBP3) hairpin sequence was identified from two
of the three pools of DNA derived from pool 66. LTBP3 was up-regulated 2.5 and 1.4
fold (p-value 9.41x10-8 and 8.18x10-5 respectively) upon growth arrest, which was
reversed when growth arrest was overcome (Table 5.5). LTBP3 was also slightly upregulated upon quiescence.
LTBPs are secreted proteins that were initially identified through their binding to the
growth factor. Three of the four known LTBPs are able to associate covalently with the
261
small latent forms of TGFβ and may be involved in their assembly, secretion and
targeting (Oklu and Hesketh 2000). LTBP3 in particular has been found to play an
essential role in the secretion and targeting of TGF-beta1 (Penttinen, Saharinen et al.
2002).
This is not in agreement with the microarray data where TGFB1 was slightly downregulated by -0.45 and -0.16 log2 fold change (p-values of 5.2x10 -4 and 2.86x10-1
respectively) upon senescence and up-regulated upon rescue by LT abrogating the p53
pathway in the CL3 EcoR cells. This suggests that although LTBP3 was reported to help
secretion of TGFB1, it might not act at the transcription level.
Interestingly, LTPB1 and LTBP2 are also up-regulated upon growth arrest indicating
that most of the LTBPs complexes are activated in the senescent cells and thereby may
result in increased TGFβ secretion upon senescence. At the transcription level, TGFB2
and TGFB3 are both up-regulated upon growth arrest by an average of 0.5 and 0.25 log 2
respectively which suggest that regulation of these two other transforming growth
factors follows a different mechanism than TGFB1.
LTBPs have subsequently been found to associate with the extracellular matrix. The
close identity between LTBPs and members of the fibrillin family, mutations in which
have been linked directly to Marfan's syndrome, suggests that anomalous expression of
LTBPs may be associated with disease. The implication of TGFβ in such a wide range
of biological responses suggests that it plays important roles in many normal cellular
functions. Consistent with these multiple roles, anomalous regulation of TGFβ activity
has been associated with the development of a number of diseases, most notably several
forms of cancer (Kimchi, Wang et al. 1988).
Studies indicated that modulation of LTBP function, and hence of TGFβ activity, was
associated with a variety of cancers (Oklu and Hesketh 2000). The contribution of
transforming growth factor (TGF) beta to breast cancer as a regulator of cancer
suppression and progression has been studied from a myriad of perspectives since
262
seminal studies more than two decades ago (Silberstein and Daniel 1987) and now
exceeds a thousand papers.
It is now generally agreed that during early tumour outgrowth, elevated TGFβ levels
suppress tumour formation (Massague 1990), whereas at later stages there is a switch
towards malignant conversion and progression. Inactivation of tumour suppressor genes,
the sequential acquisition of oncogenic mutations, and epigenetic changes within the
cancer genome divert the canonical growth inhibitory arm of the TGFβ signalling
pathway towards behaviours that increase motility, invasion and metastasis (Derynck,
Akhurst et al. 2001).
Thus, if these LTBP proteins play critical roles in controlling and directing the activity
of TGFβs, it could suggest an indirect implication in the suppression and/or the
development of cancer.
Since silencing of LTBP3 can bypass cellular senescence in CL3EcoR cells (Figure 5.8), it
suggests that LTBP3 is definitely linked with the control of cell growth and may be
playing a role in suppressing tumour progression. This is in accordance with the
identification of TGF as the cellular senescence-inducing factor in the human lung
adenocarcinoma cell line A549 (Katakura 2006). It is also in accordance with several
other reports suggesting that TGFβ1 is capable of inducing cellular senescence. For
instance, stimulation of human diploid fibroblasts with TGFβ1 triggers the appearance
of biomarkers of cellular senescence such as SA-β-Gal activity and increases steady
state mRNA levels of senescence associated genes including Apo J, fibronectin, and
M22 (Frippiat, Chen et al. 2001; Frippiat, Dewelle et al. 2002; Debacq, Heraud et al.
2005).
It is important to note that only one shRNA constructs was available at the time of the
experiment from the Open Biosystems library. Although the experiment was
successfully repeated twice with similar results, it would be valuable to use an
263
alternative silencing construct for this gene to prove that the biological consequences are
not due to an off-target effect.
5.3.4.3 ATXN10
The hair pin targeting ataxin 10 (ATXN10) was recovered from pool 9. This gene was
slightly up-regulated (1.3 fold, p-value 8.35x10-4) upon senescence arrest which was
reversed upon silencing of p53 and p21CIP1/WAF1/Sdi1 or ectopic expression of the
dominant negative E2F-DB protein (Table 5.5). Surprisingly, though, while
p21CIP1/WAF1/Sdi1 or p53 shRNA reverse the up-regulation to down-regulation, p53 GSE
does not have any effect on ATXN10 expression. WT LT, E1A and E7 expression also
had very little effect on the ATXN10 expression. However, it is difficult to conclude on
this data because there was only one oligo representing ATXN10.
Spinocerebellar ataxia type 10 (SCA10) is a dominantly inherited disorder characterized
by ataxia, seizures and anticipation caused by an intronic ATTCT pentanucleotide repeat
expansion. The ATXN10 gene encodes a novel protein, ataxin 10, known previously as
E46L, which is widely expressed in the brain and belongs to the family of armadillo
repeat proteins. Although clinical features of the disease are well characterized, nothing
is known so far about the affected SCA10 gene product, ATXN10. ATXN10 knock
down by RNAi has been shown recently to cause increased apoptosis in primary
cerebellar cultures, thus implicated in SCA10 pathogenesis (Marz, Probst et al. 2004;
Waragai, Nagamitsu et al. 2006).
This is in contrast to my finding that silencing of ATXN10 in HMF3A cells by four
different shRNAmirs did not cause apoptosis but promoted growth and permitted a
bypass of senescence growth arrest (Figure 5.9). This is not incompatible, the
differences in the biological phenotype are probably due to the cell context, but it
definitely underlines a regulating role of ATXN10 in cell growth.
264
5.3.4.4 LAYN
The Layilin (LAYN) shRNAmir was isolated from pool 58. LAYN was up-regulated by
2.0 fold upon senescence arrest (p-value 1.81x10-4) and this was reversed upon
abrogation of the growth arrest by inactivation of either the p53-p21 or the p16-pRb
pathways or both (Table 5.5).
Moreover, two different LAYN shRNAmirs were found to directly bypass senescence in
CL3ECoR cells (Figure 5.10).
Layilin is a widely expressed integral membrane hyaluronan receptor, originally
identified as a binding partner of talin located in membrane ruffles. Talin is responsible,
along with its adaptor proteins, for maintaining the cytoskeleton-membrane linkage by
binding to integral membrane proteins and to the cytoskeleton.
Recently, Layilin has been suggested to play a crucial role in lymphatic metastasis of
lung carcinoma A549 cells (Chen, Zhuo et al. 2008). That study found that suppression
of layilin expression by RNA interference significantly inhibited A549-cell invasion and
migration in vitro and lymphatic metastasis in vivo and thereby resulted in the increased
survival of tumour-bearing mice.
5.3.4.5 SGBT
The shRNAmir corresponding to Small glutamine-rich tetratricopeptide (SGBT) was
recovered from pool 59. SGBT expression was slightly up-regulated (1.3 fold, p-value
8.79x10-4) upon senescence growth arrest which was reversed upon abrogation of the
p53-p21 and p16-pRB pathways.
SGBT or hSGT, also known as viral protein U-binding protein (UBP), was originally
identified as a protein interacting with non-structural protein NS1 of parvovirus H-1
(Cziepluch, Lampel et al. 2000). SGBT has been reported to function as a molecular
265
chaperone that can associate with various cellular proteins such as the ubiquitous heatshock proteins cognate Hsc70 and Hsp70 (Liu, 1999), regulate their ATPase activity
(Tobaben, Thakur et al. 2001) and negatively influences their ability to refold denatured
proteins (Wu, Liu et al. 2001).
hSGT was also shown to physically interact with myostatin in yeast cells which
suggested a functional relationship between these proteins in skeletal muscle cells
(Wang, Zhang et al. 2003).
It was proposed that hSGT may act as a molecular chaperone that assists in secretion and
activation of myostatin together with other unidentified factors.
Myostatin is a member of the TGFβ superfamily and function as a negative regulator of
skeletal muscle growth (Grobet, Martin et al. 1997; McPherron and Lee 1997).
Myostatin shares all common features of TGFβ superfamily members. Like TGFβ,
myostatin is also present in serum and circulates in the blood of adult mice in a
biologically inactive form (Zimmers, Davies et al. 2002).
It has been suggested that assembly, secretion, and activation of TGFβ are regulated in
part by its interacting proteins, such as latency-associated proteins (LAPs) and latent
TGFβ-binding proteins (Koli, Saharinen et al. 2001).
Removal of LTBP is indispensable for TGFβ activation in cells as biological activity of
TGFβ in circulation is tightly controlled by their existence as latent complexes with
LTBPs. The regulation of TGFβ function by these proteins may have extremely
important biological implications. Altered expression of LTBPs has been associated with
development of human diseases such as cancer and atherosclerosis (Eklov, Funa et al.
1993; Mizoi, Ohtani et al. 1993) although the functional role of LTBPs in these diseases
is largely unclear.
266
Although the functional consequences of the observed interactions of hSGT with
myostatin remain unclear, it is conceivable that hSGT functions as LTBPs to regulate
myostatin secretion and thereby to determine its biological activity in skeletal muscle
cells.
This corroborates the fact that along with SGTB (Figure 5.10), LTBP3 silencing also
bypass senescence (Figure 5.8) and it is interesting that both SGTB and LTBP3 were
two of the five targets identified by the shRNA screen. This reinforces the hypothesis
that their functions are very similar and further validate their role in Cancer and
Senescence, highlighting the importance of TGFβ in these processes.
5.3.4.6 TAOK1, RAS4A and ARMCX2
An extra three targets from the primary screen were found to be of interest in the context
of senescence. Since these genes were not up-regulated upon senescence, they were not
silenced in the conditional system. However, according to the literature, they may be of
interest.
Taok1 was a gene identified from pool 13. TAOK1 mRNA expression does not vary
upon senescence however it is possible that its protein activity could be subject to
variation upon senescence.
 TAOK1
TAOK1 is known to activate the p38 MAP kinase pathway through the specific
phosphorylation of MKK3. The p38 MAPK pathway is a complex pathway responsive
to stress stimuli and involved in cell differentiation and apoptosis which has shown to
have an important causative role in senescence. It is well known that oncogenic Ras, the
constitutively active form of Ras, contributes to transformation-associated phenotypes in
immortalized cells but senescence in normal cells (Katz and McCormick 1997; Serrano,
Lin et al. 1997). It was shown that among the divergent downstream pathways of Ras,
267
the Erk MAPK pathway is responsible for the Ras-induced senescence (Lin, Barradas et
al. 1998; Zhu, Woods et al. 1998). It was also reported that p38 MAPK activation is
involved in this Ras-Erk MAPK-induced senescence (Wang, Chen et al. 2002).
This suggests that TAOK1 might be necessary to the activation of the p38 MAPK
pathway which itself play a causative role in senescence.
TAOK1 has also been identified in a genomic screen to identify human kinases and
phosphatases important for the regulation of mitotic progression. TAOK1 is a microtubule affinity-regulating kinase which is required for both chromosome congression
and checkpoint-induced anaphase delay (Draviam, Stegmeier et al. 2007). It is known to
interact with BUB1. The consequences of this interaction are not known. Interestingly,
TAOK1 over-expression has been shown in breast cancer lines (Kao, Salari et al. 2009)
and somatic mutation of TAOK1 has been described in several human cancer tissues
including glioblastoma (Parsons, Jones et al. 2008), lung cancer (Davies, Hunter et al.
2005) and digestive tract squamous cell carcinoma.
 RAS4A
This gene encodes a member of the GAP1 family of GTPase-activating proteins that has
been identified to suppress the Ras/MAPK pathway in response to an elevation of Ca2+.
Stimuli that increase intracellular Ca2+ levels result in the translocation of this protein to
the plasma membrane, where it activates Ras GTPase activity. Consequently, Ras is
converted from the active GTP-bound state to the inactive GDP-bound state and no
longer activates the downstream pathways that regulate gene expression, cell growth,
and differentiation (Lockyer, Kupzig et al. 2001). RAS4A is not up-regulated upon
senescence but its expression might be essential to some pathways that take place during
senescence and that might be necessary to the trigger or the maintenance of the cell
cycle arrest. Interestingly, K- ras4A was found to be over-expressed 2-to 3-fold higher in
lung tumour cell lines (Wang and You 2001).
268
 ARMCX2 or ALEX2
This gene encodes a member of the ALEX family of proteins and may play a role in
tumour
suppression.
The
encoded
protein
contains
a
potential
N-terminal
transmembrane domain and a single Armadillo (arm) repeat. Other proteins containing
the arm repeat are involved in development, maintenance of tissue integrity, and
tumourigenesis. The genes encoding ALEX1, ALEX2 and ALEX3 co-localize to the
same region in Xq21.33-q22.2. ALEX1 and Expression of ALEX1 and ALEX2 mRNA
was found to be lost or significantly reduced in human lung, prostate, colon, pancreas,
and ovarian carcinomas and also in cell lines established from different human
carcinomas. These genes are, however, normally expressed in cell lines derived from
other types of tumours, e.g., sarcomas, neuroblastomas, and gliomas. ALEX gene was
suggested to play a role in suppression of tumours originating from epithelial tissue, i.e.,
carcinomas (Kurochkin, Yonemitsu et al. 2001).
ARMCX2 was up-regulated upon senescence when looking at the raw microarray data;
it did not appear in the differential data set since it was also up-regulated in the heat
shock control. Since it was found amongst the targets in the primary screen, it suggests
that it may also potentially have a role in senescence which is in accordance with its
down-regulation in some tumours.
269
6
ROLE OF MICRO-RNAS IN CELLULAR SENESCENCE
6.1 SENESCENCE SPECIFIC MICRO-RNA DIFFERENTIAL EXPRESSION
6.1.1 Objectives
Micro-RNAs have recently emerged as key regulators of gene expression in many
developmental and cancer processes like cell proliferation, differentiation, cell cycle,
apoptosis and metastasis. It is actually hypothesized that probably every cellular process
is regulated at least partially by micro-RNAs, and an aberrant micro-RNA expression
signature can be the hallmark of several diseases, including cancer (Iorio and Croce
2009).
An increasing number of studies have then demonstrated that micro-RNAs can function
as potential oncogenes or tumour suppressor genes, depending on the cellular context
and on the target genes they regulate. The Aim of this chapter was to analyse
senescence-specific micro-RNAs expression in the HMF3AEcoR system in a similar way
to the one used for the genome wide microarray in order to investigate the involvement
of miRs in senescence and their potential as a tool to understand better the mechanism
behind senescence pathways.
6.1.2 Background to micro-RNA Expression Profiling Technology
Genome-wide microarray gene expression analysis has been widely utilised to
investigate human cancers and allowed the identification of important genes for both
prognostic and therapeutics (Martin, Graner et al. 2001; Chung, Sung et al. 2002; Mohr,
Leikauf et al. 2002; van 't Veer, Dai et al. 2002; Ramaswamy and Perou 2003).
Recently, microarray analysis has been enriched by the development of platforms for the
analysis of micro-RNA (miRNA) expression (Calin, Sevignani et al. 2004; Liu, Calin et
270
al. 2004). Investigation of over-expression and down-regulation of miRNAs in
senescence would represent an innovative and efficient approach from a completely
different perspective to study the senescence mechanisms. Previous studies have
demonstrated that there is a large number of deregulated miRNAs in human breast
cancer. Different miRNA expression signatures (Iorio, Ferracin et al. 2005; MertensTalcott, Chintharlapalli et al. 2007; Ma and Weinberg 2008) have also been correlated
with different prognostic parameters such as tumour size, nodal involvement, vascular
invasiveness and chemotherapy resistance (Yan, Zhou et al. 2009; Zhao, Yang et al.
2009). Here, using micro-RNA expression profiling within the HMF3AEcoR cells could
bring new answers to the understanding of cellular senescence.
One of the first microarrays to become available and used by the Massague lab was from
LC Sciences (Tavazoie, Alarcon et al. 2008) and represented a human genome-wide
miRNA array, based upon the latest release (10.1) from the Sanger miRBase Sequence
Database (catalogue number MRA-1001) and corresponds to 723 unique mature miRNA
probes.
6.1.3 HMF3AEcoR: miRNA expression profiling experimental design
My aim was to identify any significant changes in miRNA expression between
HMF3AEcoR cells incubated at 34°C and HMF3AEcoR cells incubated at 38°C for 7 days,
and to identify changes that occur due to quiescence (Figure 6.1). The quiescence
specific signal was determined by comparing the signal from HMF3AEcoR cells
incubated at 34°C with serum starved (0.5%FCS) HMF3AEcoR cells incubated at 34°C
for a week.
Each chip contained multiple redundant miRNA probe regions required to detect
miRNA transcripts (www.sanger.ac.uk/Software/Rfam/mirna/) listed in Sanger miRBase
Release 10.1. Furthermore, multiple control probes were also present on each chip as
quality controls for production, sample labelling and assay conditions.
Because of the possibility of sample variation, biological triplicates were used.
271
HMF3A
quiescent
HMF3A at
34 C
Quiescence
specific gene
expression
changes
HMF3A at
38 C
Growth arrest
specific gene
expression
changes
GA - Q
Senescence
specific gene
expression
changes
Figure 6.1: microRNAs microarray profiling strategy
Significant changes in miRNA expression between HMF3AEcoR cells incubated at 34°C and HMF3AEcoR
cells incubated at 38°C for 7 days and changes that occur due to quiescence were identified. The
quiescence specific signal was determined by comparing the signal from HMF3A EcoR cells incubated at
34°C with serum starved (0.5%FCS) HMF3AEcoR cells incubated at 34°C for a week. To obtain
senescence specific changes, the growth arrest changes occurring also in quiescence were removed.
272
Dual hybridization was set-up to make pairwise comparison of the samples as seen in
Table 6.1 to minimize inter chip errors and provide more reliable data. The triplicate
samples were also labelled with Cy3 and Cy5 alternatively to normalize the differences
in the dye incorporation.
6.1.4 Quality Control of RNA Samples
The 260 nm/230 nm and 260 nm/280 nm ratio of each extracted RNA sample were
analyzed by Nanodrop. The samples all had a 260/280 ratio above 1.8 which means that
the samples were not contaminated by proteins. The 260/230 ratio was above 1.8 for
approximately 50% of the samples leaving the other 50% of the samples with a slight
trace of ethanol contamination present in the samples.
6.1.5
miRNAs senescence specific differential expression
This analysis was designed to identify senescence-specific miR expression by
determining which miRs are differentially expressed upon the shift from 34°C to 38°C
but do not change significantly upon quiescence. Because I found that many of the upregulated changes and particularly the NF-κB targets were also up-regulated by
quiescence, my strategy was to take the quiescence expression in consideration but not
eliminate the genes also modulated by quiescence from the results. That way, I had all
the information necessary to choose targets.
Due to the prices of the arrays, it was impossible to incorporate the HMF3S at 34°
versus 38°C to eliminate changes due to the temperature shift.
273
Table 6.1: Dual-hybridization analysis
Dual hybridization was set-up to make pairwise comparison of the samples and to minimize inter chip
errors and provide more reliable data. The triplicate samples were also labelled with Cy3 and Cy5
alternatively to normalize the differences in the dyes incorporation.
274
In the first step, the genes differentially expressed upon growth arrest were identified
after filtering for low signals (at least > 32). Expression results for the 86 remaining
miRs for growth arrest and 64 for quiescence after filtering for low signal and for
significant results (p-value at least < 0.1) are displayed in table 6.2 and 6.3.
The Figure 6.2 represents the heat map of the 86 micro-RNAs differentially expressed
with a mean p-value < 0.1 upon growth arrest and the 64 micro-RNAs differentially
expressed with a mean p-value < 0.1 upon quiescence. This cut off value was suggested
by LC Sciences for significant samples in this analysis even though it is much higher
than the one used for the Affymetrix data.
The first three columns and the three last columns of each heat map represent the
triplicate samples at 34°C against samples at 38°C. It is possible to note that the
triplicate samples displayed the same colour changes upon growth arrest or quiescence
which validates the reproducibility of the data even though column 2 shows slight
inconsistency with the other two.
The expression changes upon both growth arrest and quiescence for the 86 miRs with a
mean p-value < 0.1 upon growth arrest were overlapped and the difference between
growth arrest and quiescence differential was calculated.
Micro-RNAs were considered specifically differential upon growth arrest only if the
difference between growth arrest and quiescence differentials was > 1 or <-1 in log2 fold
change (which corresponds to a two-fold difference in expression). In addition, the list
was also filtered for the miRs, which exhibited changes in expression at least >0.5 or <0.5 log2 fold change upon growth arrest.
This gave a micro-RNA list of 33 micro-RNAs of which 18 were up-regulated upon
growth arrest and 15 were down regulated.
The up and down-regulated micro-RNAs and their respective expression levels are
shown in Tables 6.4 and 6.5.
275
Group
34
No.
197
289
26
123
127
237
152
396
697
130
320
101
64
617
125
112
144
351
62
121
256
201
16
223
70
650
163
246
406
276
373
175
274
117
19
643
14
700
30
Reporter Name
hsa-miR-20a
hsa-miR-320
hsa-miR-106a
hsa-miR-15a
hsa-miR-16
hsa-miR-25
hsa-miR-18a
hsa-miR-455-3p
hsa-miR-92a
hsa-miR-17
hsa-miR-34a
hsa-miR-146a
hsa-miR-128
hsa-miR-638
hsa-miR-15b
hsa-miR-149*
hsa-miR-185
hsa-miR-376c
hsa-miR-127-3p
hsa-miR-155
hsa-miR-29a
hsa-miR-21
hsa-let-7i
hsa-miR-221
hsa-miR-130b
hsa-miR-708
hsa-miR-193a-5p
hsa-miR-27b
hsa-miR-487b
hsa-miR-30b
hsa-miR-423-5p
hsa-miR-199a-3p
hsa-miR-30a
hsa-miR-152
hsa-miR-100
hsa-miR-663
hsa-let-7g
hsa-miR-92b
hsa-miR-107
p-value
5.49E-04
1.17E-03
1.27E-03
2.50E-03
3.29E-03
4.46E-03
4.91E-03
5.88E-03
6.11E-03
8.49E-03
8.54E-03
1.13E-02
1.70E-02
1.84E-02
1.89E-02
1.92E-02
1.93E-02
1.99E-02
2.18E-02
2.23E-02
2.37E-02
2.61E-02
2.64E-02
2.64E-02
3.03E-02
3.07E-02
3.66E-02
3.68E-02
4.10E-02
4.16E-02
4.45E-02
5.29E-02
6.32E-02
6.34E-02
8.02E-02
8.87E-02
9.04E-02
9.11E-02
9.36E-02
Mean
2,566
3,742
1,605
756
13,974
3,615
507
364
7,050
1,932
55
83
582
2,531
7,796
256
585
115
293
4,558
15,363
41,589
12,627
5,028
410
246
1,021
4,034
380
472
1,862
6,741
310
428
6,956
704
3,999
2,173
1,277
276
Log2
(38/34
)
Group
38
StDev
203
81
164
166
1,466
390
42
67
1,183
366
21
11
52
496
3,119
5
91
16
97
1,492
2,653
3,185
742
1,373
98
64
124
713
57
42
379
557
90
76
1,647
375
1,374
1,211
333
Mean
925
5,107
703
167
6,359
1,705
109
940
3,292
879
723
1,676
360
5,921
1,469
537
959
492
990
1,475
29,875
21,404
9,856
15,117
184
589
1,492
2,491
565
376
1,205
9,064
585
655
11,677
1,491
2,144
788
2,025
StDev
76
196
43
28
716
207
19
137
306
99
112
871
52
1,287
58
91
105
168
49
470
314
3,554
889
2,250
47
18
81
401
75
27
124
1,287
26
117
1,310
380
462
69
396
-1.47
0.45
-1.19
-2.18
-1.14
-1.08
-2.22
1.37
-1.10
-1.14
3.71
4.33
-0.69
1.23
-2.41
1.07
0.71
2.10
1.76
-1.63
0.96
-0.96
-0.36
1.59
-1.16
1.26
0.55
-0.70
0.57
-0.33
-0.63
0.43
0.92
0.62
0.75
1.08
-0.90
-1.46
0.66
285 hsa-miR-31
Following transcripts
231 hsa-miR-23a*
199 hsa-miR-20b
135 hsa-miR-181b
258 hsa-miR-29b
129 hsa-miR-16-2*
247 hsa-miR-27b*
497 hsa-miR-532-5p
555 hsa-miR-584
401 hsa-miR-485-3p
400 hsa-miR-484
238 hsa-miR-25*
133 hsa-miR-181a*
370 hsa-miR-421
297 hsa-miR-329
418 hsa-miR-495
154 hsa-miR-18b
542 hsa-miR-574-3p
544 hsa-miR-575
379 hsa-miR-431
119 hsa-miR-154
415 hsa-miR-493
305 hsa-miR-337-5p
404 hsa-miR-486-5p
286 hsa-miR-31*
205 hsa-miR-212
203 hsa-miR-210
46 hsa-miR-1229
100 hsa-miR-145*
72 hsa-miR-132
84 hsa-miR-138
109 hsa-miR-148b
55 hsa-miR-125a-3p
229 hsa-miR-224
329 hsa-miR-362-5p
76 hsa-miR-134
375 hsa-miR-424*
633 hsa-miR-654-3p
139 hsa-miR-182
381 hsa-miR-432
300 hsa-miR-331-3p
316 hsa-miR-342-3p
9.73E-02 4,350 1,582 8,595
680
0.98
are statistically significant but have low signals (s
1.69E-05
488
10
136
5 -1.84
1.38E-03
409
76
85
11 -2.27
1.47E-03
437
27
213
20 -1.04
3.69E-03
32
5
146
8
2.18
6.88E-03
29
3
14
2 -1.02
8.93E-03
180
48
37
2 -2.27
1.09E-02
56
6
106
17
0.93
1.24E-02
29
2
84
15
1.55
1.28E-02
105
9
165
19
0.65
1.66E-02
159
19
98
13 -0.69
1.92E-02
121
57
30
8 -2.03
2.05E-02
30
4
19
2 -0.69
2.46E-02
62
14
28
7 -1.12
2.64E-02
35
7
69
3
0.97
2.74E-02
56
18
161
14
1.52
2.79E-02
101
6
42
11 -1.26
2.86E-02
60
8
180
56
1.59
3.28E-02
40
4
119
35
1.57
3.51E-02
28
6
52
10
0.87
3.52E-02
35
8
78
10
1.14
3.70E-02
22
4
37
6
0.80
3.82E-02
30
10
72
18
1.25
4.01E-02
18
4
31
6
0.80
4.01E-02
18
4
38
12
1.12
4.20E-02
35
6
24
2 -0.54
4.47E-02
47
3
123
40
1.40
4.49E-02
29
4
22
2 -0.44
5.19E-02
38
10
20
4 -0.95
5.40E-02
283
117
100
16 -1.50
5.60E-02
82
18
145
34
0.83
5.81E-02
31
2
26
2 -0.28
5.86E-02
40
4
24
5 -0.74
5.87E-02
135
46
65
17 -1.07
6.07E-02
37
2
43
2
0.19
6.22E-02
384
34
302
34 -0.35
6.90E-02
129
52
50
4 -1.38
7.05E-02
40
2
84
27
1.06
7.09E-02
29
8
47
9
0.72
8.06E-02
130
56
264
11
1.02
8.20E-02
48
6
91
28
0.92
8.82E-02
58
18
33
6 -0.83
Table 6.2: Raw microarray results for growth arrest
Results microarray for the micro-RNAs transcripts with significant results (p-value > 0.1)
277
Group 34
Group Q
Log2
(Q/34)
No. Reporter Name
p-value
Mean
StDev
Mean
StDev
617 hsa-miR-638
1.64E-04
2,102
147
6,888
367
1.71
101 hsa-miR-146a
2.36E-03
176
29
1,692
139
3.27
99 hsa-miR-145
4.10E-03
737
168
206
28
-1.84
246 hsa-miR-27b
5.07E-03
4,862
435
2,252
80
-1.11
117 hsa-miR-152
5.40E-03
840
74
508
43
-0.73
239 hsa-miR-26a
6.38E-03
8,256
271
9,925
332
0.27
130 hsa-miR-17
6.74E-03
1,622
129
1,118
61
-0.54
26 hsa-miR-106a
7.35E-03
1,509
117
1,051
62
-0.52
112 hsa-miR-149*
7.96E-03
312
101
1,363
312
2.13
16 hsa-let-7i
1.20E-02
12,578
82
10,889
295
-0.21
695 hsa-miR-923
1.24E-02
3,289
352
5,297
581
0.69
121 hsa-miR-155
1.38E-02
5,525
1,070
2,745
390
-1.01
197 hsa-miR-20a
1.62E-02
2,480
285
1,373
81
-0.85
135 hsa-miR-181b
1.64E-02
393
96
184
23
-1.10
95 hsa-miR-143
2.11E-02
625
229
185
41
-1.76
278 hsa-miR-30c
2.13E-02
1,248
109
1,746
25
0.48
234 hsa-miR-24
2.92E-02
6,275
1,048
3,507
633
-0.84
720 hsa-miR-99a
3.06E-02
1,522
216
2,359
314
0.63
123 hsa-miR-15a
3.38E-02
801
123
1,412
288
0.82
258 hsa-miR-29b
3.39E-02
71
34
482
278
2.76
244 hsa-miR-27a
4.25E-02
6,381
1,319
3,498
178
-0.87
643 hsa-miR-663
4.62E-02
588
299
2,069
266
1.81
199 hsa-miR-20b
6.04E-02
654
67
479
79
-0.45
232 hsa-miR-23b
6.70E-02
16,932
1,069
13,987
1,407
-0.28
223 hsa-miR-221
6.73E-02
5,695
1,882
10,960
700
0.94
127 hsa-miR-16
8.05E-02
11,327
1,359
14,836
1,851
0.39
276 hsa-miR-30b
8.52E-02
1,040
185
1,508
292
0.54
116 hsa-miR-151-5p
9.26E-02
2,192
202
1,432
328
-0.61
225 hsa-miR-222
9.78E-02
7,788
2,027
12,033
408
0.63
Following transcripts are statistically significant but have low signals (signal < 500
401 hsa-miR-485-3p
1.77E-03
58
5
26
3
-1.16
100 hsa-miR-145*
2.62E-03
43
4
17
2
-1.35
247 hsa-miR-27b*
3.65E-03
117
16
44
6
-1.41
418 hsa-miR-495
6.60E-03
108
15
47
7
-1.20
357 hsa-miR-379*
8.13E-03
42
5
20
1
-1.10
224 hsa-miR-221*
1.06E-02
125
14
235
36
0.92
544 hsa-miR-575
1.10E-02
40
9
144
11
1.84
173 hsa-miR-197
1.12E-02
129
6
60
8
-1.09
368 hsa-miR-411*
1.50E-02
65
12
27
2
-1.28
8 hsa-let-7d*
1.74E-02
55
2
32
4
-0.79
278
76
104
152
399
29
229
318
149
658
406
231
55
134
255
297
361
375
431
649
381
348
47
60
356
hsa-miR-134
hsa-miR-146b-5p
hsa-miR-18a
hsa-miR-483-5p
hsa-miR-106b*
hsa-miR-224
hsa-miR-345
hsa-miR-187*
hsa-miR-765
hsa-miR-487b
hsa-miR-23a*
hsa-miR-125a-3p
hsa-miR-181a-2*
hsa-miR-299-5p
hsa-miR-329
hsa-miR-382
hsa-miR-424*
hsa-miR-503
hsa-miR-7
hsa-miR-432
hsa-miR-376a
hsa-miR-1231
hsa-miR-126
hsa-miR-379
1.85E-02
2.20E-02
2.34E-02
2.66E-02
2.80E-02
2.99E-02
3.04E-02
3.46E-02
3.55E-02
3.83E-02
4.00E-02
4.51E-02
4.62E-02
5.18E-02
5.83E-02
6.21E-02
6.36E-02
6.66E-02
7.71E-02
7.78E-02
8.21E-02
8.73E-02
9.18E-02
9.72E-02
189
44
277
26
22
313
31
28
18
256
272
21
71
212
63
355
103
55
90
84
41
23
34
433
24
12
53
15
2
73
2
18
4
59
36
2
1
45
16
91
17
9
36
24
9
6
8
55
105
146
131
108
33
134
38
169
35
131
119
29
41
126
34
193
62
37
37
46
25
74
66
235
Table 6.3: Raw microarray results for quiescence
Results microarray for the micro-RNAs transcripts with significant results (p-value > 0.1)
279
17
59
30
36
5
12
3
70
3
30
31
3
9
21
3
48
17
7
17
10
7
34
24
79
-0.84
1.72
-1.08
2.03
0.60
-1.23
0.31
2.59
0.96
-0.96
-1.20
0.44
-0.80
-0.74
-0.87
-0.88
-0.73
-0.59
-1.26
-0.88
-0.70
1.66
0.96
-0.88
34°C
samples
38°C
samples
A
34°C
Quiescent
samples samples
B
Mean p-value < 0.1
Figure 6.2: Differential microRNAs upon growth arrest (A) and quiescence (B)
Heat map of the 86 microRNAs differentially expressed with a mean p-value < 0.1 upon growth arrest (A)
and the 64 microRNAs differentially expressed with a mean p-value < 0.1 upon quiescence (B). The first
three columns and the three last columns of each heat map represent the triplicate samples at 34°C against
samples at 38°C.
280
Reporter Name
LogFC GA
p-value
LogFC
quiescence
p-value
Difference
logFC
hsa-miR-146a
4.33
1.13E-02
3.27
2.36E-03
1.06
hsa-miR-34a
3.71
8.54E-03
-0.54
3.62E-01
4.25
hsa-miR-376c
2.10
1.99E-02
-1.05
1.34E-01
3.15
hsa-miR-127-3p
1.76
2.18E-02
-0.20
4.87E-01
1.96
hsa-miR-574-3p
1.59
2.86E-02
0.05
9.11E-01
1.53
hsa-miR-495
1.52
2.74E-02
-1.20
6.60E-03
2.72
hsa-miR-210
1.40
4.47E-02
0.26
3.75E-01
1.13
hsa-miR-455-3p
1.37
5.88E-03
0.02
9.58E-01
1.35
hsa-miR-708
1.26
3.07E-02
-0.09
9.02E-01
1.35
hsa-miR-154
1.14
3.52E-02
0.14
6.07E-01
1.00
hsa-miR-149*
1.07
1.92E-02
2.13
7.96E-03
-1.06
hsa-miR-654-3p
1.06
7.05E-02
-0.23
5.47E-01
1.28
hsa-miR-432
1.02
8.06E-02
-0.88
7.78E-02
1.91
hsa-miR-329
0.97
2.64E-02
-0.87
5.83E-02
1.84
hsa-miR-185
0.71
1.93E-02
-0.35
3.45E-01
1.06
hsa-miR-485-3p
0.65
1.28E-02
-1.16
1.77E-03
1.82
hsa-miR-152
0.62
6.34E-02
-0.73
5.40E-03
1.34
hsa-miR-487b
0.57
4.10E-02
-0.96
3.83E-02
1.54
The samples considered not significant (p -value>0.1) are shown in yellow
Table 6.4: Up-regulated micro-RNAs upon senescence
18 microRNAs were up-regulated upon growth arrest and this table represents their respective expression
levels upon growth arrest and quiescence
281
LogFC
p-value
quiescence
-0.10
5.94E-01
Difference
logFC
-2.30
Reporter Name
LogFC GA
p-value
hsa-miR-15b
-2.41
1.89E-02
hsa-miR-20b
-2.27
1.38E-03
-0.45
6.04E-02
-1.82
hsa-miR-18a
-2.22
4.91E-03
-1.08
2.34E-02
-1.14
hsa-miR-15a
-2.18
2.50E-03
0.82
3.38E-02
-3.00
hsa-miR-29b-1*
-2.13
9.30E-02
-0.34
3.96E-01
-1.79
hsa-miR-25*
-2.03
1.92E-02
0.26
4.19E-01
-2.30
hsa-miR-132
-1.50
5.40E-02
-0.31
5.30E-01
-1.18
hsa-miR-15b*
-1.36
9.64E-02
0.43
5.16E-01
-1.79
hsa-miR-130b
-1.16
3.03E-02
0.02
9.51E-01
-1.17
hsa-miR-16
-1.14
3.29E-03
0.39
8.05E-02
-1.53
hsa-miR-16-2*
-1.02
6.88E-03
0.15
9.25E-01
-1.17
hsa-miR-195
-1.00
9.75E-02
1.51
3.53E-01
-2.51
hsa-miR-193b*
-0.82
9.54E-02
0.99
3.81E-01
-1.81
hsa-miR-125a-3p
-0.74
5.86E-02
0.44
4.51E-02
-1.18
hsa-miR-181a*
-0.69
2.05E-02
0.34
4.94E-01
-1.03
The samples considered not significant (p -value>0.1) are shown in yellow
Table 6.5: Down-regulated microRNAs upon senescence
15 microRNAs were up-regulated upon growth arrest and this table represents their respective expression
levels upon growth arrest and quiescence
282
6.1.6 Up-regulated micro-RNAs
The microarray analysis permitted the designation of 18 micro-RNAs specifically upregulated upon senescence; Table 6.4 shows them ranked by fold change for growth
arrest. Some of the miRs were also up-regulated upon quiescence (miR-146a: 3.27 log2
fold change- highest up-regulated) whereas others are down-regulated (miR-495: -1.2
log2 fold change- highest down-regulated). It was particularly interesting that 34a was
amongst the up-regulated miRs, exactly the way it had been described extensively in the
last few years. MiR-34a has also been previously linked to cancer, apoptosis and growth
arrest.
6.1.7 Down-regulated micro-RNAs
The microarray analysis permitted the designation of 15 micro-RNAs down-regulated
upon senescence growth arrest some of which were also down-regulated upon
quiescence (miR-18a: 1.08 log2 fold change- highest down-regulated) whereas others
were up-regulated (miR-195: 1.5 log2 fold change- highest up-regulated) (table 6.5).
MiR-372 and MiR-373 were absent from the chip I used which explain why they are
absent from the differential results.
6.2 BIOLOGICAL VALIDATION BY GROWTH COMPLEMENTATION
ASSAY IN THE HMF3A CELLS
6.2.1 Objectives
The objectives of the biological validation were to confirm the involvement of the
differentially expressed micro-RNAs in the senescence process and more specifically the
down-regulated micro-RNAs. To address this issue, it was important to investigate
whether the down regulation was essential to triggering the growth arrest or only a
consequence. In theory, if the down-regulation of a miR was causal to senescence, its
283
up-regulation by ectopic expression should reverse senescence and not allow the growth
arrest. The complementation using the CL3EcoR model was a great way to assess this
biologically. Ultimately this should be also validated in primary cells. Since the LC
Science chips were very expensive, I decided that it would be easier and more cost
effective to simply validate them by ectopic expression since the miR-Vec were
available.
6.2.2 Validation by ectopic expression
6.2.2.1 miR Vec clones
Agami and colleagues at NKI created a library of miRs cloned into a retroviral
expression vector under the control of a CMV promoter (Voorhoeve, le Sage et al.
2006). Geneservice provides these clones as a micro-RNA library (MiR-Lib). The
following miRs were purchased and used for the complementation assay and
information about each miR was extracted from miRbase (www.mirbase.org):
Mir-186*: hsa-miR-186: A light dye bias has been found for this miR on the microarray
which has to be taken into consideration as well as the p-value which was not within the
filtering threshold applied which is why that micro-RNA was not in the Table 6.5.
Nevertheless, since it was the most highly down-regulated micro-RNA, it was still
selected for ectopic expression. This micro-RNA is also located within intron 8 of the
ZRANB2 gene and does not form part of a miR cluster.
MiR-15b, miR-15b*, mir-15a and miR-16: hsa-miR-15a and hsa-miR-15a/16: MiR-15 a,
miR-15b and miR-16 are clustered together. There were only two clones available:
miRVec hsa-miR-15a/16 or miRVec hsa-miR-15a. MiR-15a 16-1 cluster is found in
the intron of a well-defined non-coding RNA gene, DLEU2.
284
MiR-20b: hsa-miR-20b: This is clustered with hsa-miR-106a, hsa-miR-18b, hsa-miR19b-2, hsa-miR-92a-2 and hsa-miR-363. There are no overlapping transcripts with this
miR.
MiR-18a: hsa-miR-18a: This miR is expressed as part of a cluster of intronic RNAs,
including miR-17, miR-18a, miR-19a, miR-20a, miR-19b-1 and miR-92b Due to the
cloning method used to generate the miR-Vec clones and the close proximity within the
cluster; miR-Vec hsa-miR-18a contains both miRs hsa-miR-17a and hsa-miR-19a in
addition to hsa-miR-18a.
MiR-29b-1*: hsa-miR-29b: It is clustered with miR-29a and is located within the
transcript AC016831.7 within intron 2.
MiR-25*: hsa-miR-25: This miR is clustered with miR-93 and 106b. The miR 106b-25
cluster consists of these three miRNAs and is located in the 13 th intron of MCM7
MiR-132: Not available
MiR-130b: hsa-miR-130b: It is clustered with miR-301b. It overlaps with two transcripts
in their intronic region: PPIL2 with it sense sequence and TOP3B with its antisense
sequence.
MiR-195: hsa-miR-195: It overlaps with transcript AC027763.1 within intron 1. It is
clustered with miR-497.
Mir-193b: hsa-miR-193b: Clustered with miR-365-1. It does not overlap with any
transcript.
MiR-125a and miR-181a: These clones would not grow. After two attempts, they were
put aside.
285
hsa-let-7a1 was also obtained as a negative control as its expression does not vary upon
senescence or quiescence. Hsa-miR-7a1 miR-Vec clone contains Hsa-miR-let-7a1, HsamiR-let-7a2 and Hsa-miR-let-7a3. All three sequences correspond to the same identical
Let-7a* mature sequence. Let-7a does not overlap with any transcript and is clustered
with Let-7-f-1 and Let-7-d.
Hsa-miR-373 and hsa-miR-372 clones were also obtained as controls since Voorhoeve
et al (2006) have shown that they are able to immortalise BJ cells and overcome RAS
induced premature senescence in conjunction with hTERT. Hsa-miR-372 miR-Vec
clone contains both Hsa-miR-372 and Hsa-miR-371. Unfortunately, the microarray did
not provide any information on these two miRs, probably because their expression signal
was too weak.
Some additional potential candidates were also identified: miR-92b, miR-218, miR-128,
miR-423-5p and Let-7g were selected for the complementation assay. These 5 microRNAs were originally identified in my first analysis of the micro-RNAs microarray and,
although I later realised that this first analysis was wrong and produced the one
described in this chapter, these micro-RNAs were obtained and tested.
MiR-218: hsa-miR-218: This miR overlaps with SLIT2 transcript within the intron 15. It
isn‘t clustered.
MiR-92b*: hsa-miR-92b: The Hsa-miR-92b miR-Vec clone contains only Hsa-miR-92a;
No clone available within NKI library for Hsa-miR-92b. Hsa-miR-92a-1 is clustered
with hsa-miR-17, hsa-miR-18a, hsa-miR-19a, hsa-miR-20a and hsa-miR-19b-1. This
cluster does not overlap with any transcripts.
MiR-128: hsa-miR-128 mir-Vec clone contains both Hsa-miR-128 and Hsa-miR-128b
(both corresponding to the same identical mature sequence) and miR-128 overlaps with
the intronic region of ARPP21.
286
Let-7g: hsa-miR-Let7g miR-Vec is not clustered and overlaps with the transcript
WDR82 within the intron 2.
MiR-423-5p: hsa-miR-423-5p miR-Vec clone contains both both Hsa-miR-423-5p and
Hsa-miR-423. It also shows that miR-423-5p overlaps with the transcript AC104984.3
within intron 1.
6.2.2.2 Sequencing of the MiRVec clones
The clones were sequenced to check the miR mature sequence. All clones
contained the correct mature miR sequence.
6.2.3 Complementation assay with the miR-Vec clones
Candidate miRs hsa-miR-186, hsa-miR-15a/16, hsa-miR-20b, hsa-miR-18a, hsamiR-130b, hsa-miR-92b, hsa-miR-25, hsa-miR-218, hsa-miR-195, hsa-miR193b, hsa-let-7a1, hsa-miR-373, hsa-miR-372, hsa-miR-92b, hsa-miR-218, hsamiR-128, hsa-miR-423-5p and hsa-Let-7g were packaged into amphotropic
viruses. pRS p21F RNAi constructs was used as a positive control and different
mixes of the miR-Vec clones were used as negative controls. These different viral
supernatants were applied to HMF3A cells and a selection of the infected cells
was performed. The cells were then reseeded and shifted to 38°C for 3 weeks
either in a 6-well plate format at 10,000 or 30,000 cells or in T-75 cm2 format at
1.2x105 or 0.5x105. The cells were stained after 3 weeks and the resulting
colouration was scanned and the number of colonies counted.
287
6.2.3.1 MiR-18a, miR-130b, miR-372, miR-373 and Let7a
The first step was to validate both the Let7a negative control and the miR-373 and miR372 positive controls. In addition, two other miRs selected from the microarray were
tested in this experiment: miR-18a and miR-130b.
The results show the number of colonies obtained after the same experiment was
repeated in various formats. The variety of formats permitted me to assess which one
would give the most consistent results: in 6-well plates with the cells seeded at 10,000
and 30,000 cells per well (Figure 6.3A), in 10 cm plates (data not shown) and in T-75
flasks with the cells seeded at 1.2x105 (Figure 6.3B). All formats gave similar results
although the flasks presented the less stressing option for the cells (healthier phenotype)
and was slightly better for statistics due to the higher number of cells. It is possible to
note that Let-7a was an excellent negative control and did not yield growing colonies
under any of the conditions employed. In contrast, miR-372 readily yielded densely
growing colonies. MiR-373 also abrogated growth arrest although not as efficiently.
The two other miRs tested here, namely miR-18a and miR-130b did bypass the growth
arrest but not as efficiently as the miR-372 control. Even though miR-18a and miR-130b
did rescue in every single repeat experiment, colonies were much smaller than miR-372
and 373. MiR-130b was also tested in flasks reseeded at 1.2x105 which showed the
results: a low number of small colonies (Figure 6.4A). For this reason miR-18a and
miR-130b were not pursued further.
6.2.3.2 MiR-92b, miR-15a, miR-16, miR-195 and miR-25
MiR-25, miR-15a, miR92b and miR-16 were tested alongside miR-372 and Let7a. Once
again, the format in which miR-25 was tested was varied but they all gave consistent
results: in 10 cm plates (data not shown), in 15cm plates (data not shown) and in T75
flasks seeded at both densities of 1.2x105 (Figure 6.4A) and 0.5x105 (Figure 6.4B).
288
Figure 6.3: Ectopic expression of miR-18a, miR-130b, miR-373 and miR-372
HMF3AEcoR cells were infected in triplicate with retrovirus expressing the indicated miRs expression
constructs and assayed for growth complementation at 38°C in 6-well plates (A) and in T-75cm2 flasks at
1.2x105(B). After 3 weeks the number of growing colonies was counted.
289
Figure 6.4: Ectopic expression of miR-25, miR-92b, miR-195, miR-15a and miR-16a
HMF3AEcoR cells were infected in triplicate with retrovirus expressing the indicated miRs expression
constructs and assayed for growth complementation at 38°C in T-75cm2 flasks at 1.2x10^5 (A) or at
0.5x10^5 (B). After 3 weeks the number of growing colonies was counted.
290
In each and every experiment, miR-25 gave strong rescue at levels almost comparable to
miR-372. In comparison, miR-15a and miR-16 yielded no colonies and miR-92b yielded
a low number of growing colonies (Figure 6.4A). MiR-195 produced a high number of
densely growing colonies, even higher than miR-372 and miR-25 (Figure 6.4B).
In conclusion, miR-25 and miR-195 appeared to be good miR targets to study in this
conditional cell system and were selected for further investigation. MiR-15a, miR-16
and miR-92b were not pursued further since they were unable to overcome growth
arrest. These experiments were repeated with the same results.
6.2.3.3 MiR-195, miR-218, miR-20b, miR-29b, miR-186
and miR-25
In this experiment, the following miRs were examined (Figure 6.5): miR-15a, 16 and
195 were confirmed; miR-20b, 29b, 218, and 186 were also tested. The analysis was
done in various different formats (data not shown) including T-75 flasks at 0.5x105
(Figure 6.5).
All formats gave similar results: miR-20b and miR-29b gave weak rescue and were
consequently not further pursued, miR-186, miR-25 and miR-193b yielded a low
number of colonies but very large and densely growing and therefore were selected for
further analysis and finally miR-218 and 195 gave a very strong rescue at a level above
miR-372 and therefore were also selected for further analysis.
6.2.3.4 MiR-128, miR-423-5p and Let7g
The complementation assay was performed in various formats (data not shown)
including in 6-well plates (Figure 6.6A) and in T-75 flask at the higher density (Figure
6.6B). The results show that miR-128 does not really rescue from growth arrest in a
reproducible manner (Figure 6.6A and B). For these reasons, miR-128 was dropped
from further investigation.
291
Figure 6.5: Ectopic expression of miR-29b, miR-20b, miR-186, miR-193b and miR-218
HMF3AEcoR cells were infected in triplicate with retrovirus expressing the indicated miRs expression
constructs and assayed for growth complementation at 38°C in T-75cm2 flasks at 0.5x10^5. After 3
weeks the number of growing colonies were counted.
292
Figure 6.6:Ectopic expression of miR-423-5p, miR-128 and Let-7g
HMF3AEcoR cells were infected in triplicate with retrovirus expressing the indicated miRs expression
constructs and assayed for growth complementation at 38°C in 6-well plates at 10,000 cells per well (A)
or in T-75cm2 flasks at 1.2x10^5 (B). After 3 weeks the numbers of growing colonies were counted.
293
The results show that Let-7g does not really rescue with a very low number of growing
colonies similarly to miR-128. For these reasons, Let-7g was dropped for further
investigation. Surprisingly, the results were very good for miR-423-5p which was able
to rescue the cells from senescence in all experiments. MiR-423-5p ectopic expression
was then further investigated along with miR-195, miR-218, miR-25 and miR-372 by
complementation assay in additional experiments and in different format: In T-75 flasks
seeded at 0.5x105 (Figure 6.7A) and 1x105 (Figure 6.7B) and in 10cm plates seeded at
1x105 (Figure 6.7C). This showed that miR-423-5p rescued at a level comparable to
miR-195, miR-25 and miR-218 in a reproducible manner and therefore was chosen for
further analysis.
Two extra complementation assays were performed with the positive miRs: miR-423-5p,
miR-195, miR-25, miR-218 and miR-186 in 6 well-plates (Figure 6.8A) and T-75 flasks
(Figure 6.8B). The results confirmed that miR-423-5p, miR-195 and miR-218 yield the
highest level of rescue comparable to miR-372 whereas miR-186, miR-25 and miR-193b
were less efficient but produced larger colonies. All were selected for further analysis.
6.2.4 Overlapping with the microarray data and the shRNA screen
6.2.4.1 MiR-25
A MiR-25 target gene list can be produced from the website miRportal
(http://140.116.247.50:800/miRNA/web/index.jsp)
which
integrates
micro-RNA
interacting targets from various prediction algorithms (MiRanda, TargetScan and
miRtarget), biological pathway information (Kegg, biocarta and GenMap), and microRNA literature.
294
Figure 6.7: Ectopic expression of miR-423-5p, miR-218 and miR-195
HMF3AEcoR cells were infected in triplicate with retrovirus expressing the indicated miRs expression
constructs and assayed for growth complementation at 38°C in T-75cm2 flasks at 0.5x10^5 (A) or at
1x10^5 (B) or in 10 cm plates at 1x10^5 (C). After 3 weeks the number of growing colonies was counted.
295
Figure 6.8: Ectopic expression of miR-423-5p, miR-186, miR-20, miR-193b, miR-29b, miR-25, miR218 and miR-195
HMF3AEcoR cells were infected in triplicate with retrovirus expressing the indicated miRs expression
constructs and assayed for growth complementation at 38°C in 6-well plates (A) or in T-75cm2 flasks at
0.5x10^5 (B). After 3 weeks the number of growing colonies was counted.
296
The gene list for miR-25 mRNA targets was overlapped with the results of the
microarray analysis. 69 of the differentially up-regulated genes were predicted targets of
miR-25 including BTG2 and GRAMD3 which were shown to rescue the cells in the
complementation assay in chapter 2. Two genes were both targets revealed in the
shRNA screen in chapter 3 and predictive target of miR-25, namely LUZP1 and Rab23.
Interestingly, LUZP1 was also up-regulated upon senescence.
6.2.4.2 MiR-195
The same analysis was performed for miR-195. The overlap of the mRNA target gene
list provided by the website with the microarray data revealed 75 genes that were both
predictive targets for miR-195 and also up-regulated upon senescence. This included the
genes CCNE1 and GRAMD3, also a target of miR-25 (see above). 13 genes could be
overlapped between the predictive miR-195 target gene list and the shRNA screen hit
list. Interestingly, LUZP1 was one of them.
6.2.4.3 MiR-218
The same analysis was performed for miR-218. 92 genes were both predictive target for
miR-218 and up-regulated upon senescence. This list included the genes CCNE1 which
is also a target of miR-195 and SCN2A (see Chapter 4). Only one gene namely
PPARGC1A could be overlapped between the miR-218 predictive target gene list and
the shRNA screen hit list. There was no overlap between the three lists.
6.2.4.4 MiR-193b
The same analysis was performed with miR-193b. 12 genes were both predictive target
for miR-193b and up-regulated upon senescence. Four genes namely PPARGC1A,
GRM3, KCNJ2 and WDFY2 could be overlapped between the miR-193b predictive
target gene list and the shRNA screen hit list. There was no overlap between the 3 lists.
297
6.2.4.5 MiR-186
The same analysis was performed with miR-186. 24 genes were both predictive target
for miR-186 and up-regulated upon senescence. Only three genes namely IL2, PPP4R2
and TFG could be overlapped between the miR-186 predictive target gene list and the
shRNA screen hit list. There was no overlap between the three lists.
6.2.4.6 MiR-423-5p
The miR-423-5p predicted target genes were examined. Only 10 genes were both
predicted target of mR-423-5p and up-regulated upon senescence. There was no overlap
with the shRNA screen results.
6.3 EXPRESSION PROFILING OF HMF3A CELLS IN WHICH GROWTH
ARREST WAS OVERCOME BY ECTOPIC EXPRESSION OF MIRS
6.3.1 Objectives
To dissect the role of micro-RNAs in cellular senescence, it was critical to identify what
were their downstream targets. Are they identical or different and what are their
relationships to previously identified targets with different approaches?
In order to investigate further exactly how these micro-RNAs affect gene expression and
furthermore which group of genes preferentially have their expression affected upon
expression of which micro-RNA and also in order to overlay these results with the list of
predicted target genes, a microarray profiling analysis of the mRNA from cells
expressing the various micro-RNAs which abrogated the growth arrest was designed and
performed.
298
6.3.2 Microarray Strategy
To minimise sources of technical variability, each experimental condition was analyzed
using biological triplicates. Specifically, three cultures were processed in parallel and
mRNA was extracted from each culture, as suggested by Lee and colleagues (Lee, Kuo
et al. 2000). In addition, the cultures were all derived from the same batch of HMF3A
cells. All cultures were developed by abrogating growth arrest upon miRNA expression.
To identify the changes in gene expression that occur upon ectopic expression of each
selected micro-RNA, triplicate independent biological samples of RNA extracted from
HMF3AEcoR cells growing at 38C for 7 days and expressing the negative control Let7a
or the relevant tested micro-RNAs, were analysed by expression profiling (Figure 6.9).
This data was then compared to the 8064 differential data set described in chapter 2. The
log2 FC of ―Let7a vs miR-X‖ was calculated for each of the micro-RNAs and compared
to the log2 FC upon senescence.
To identify genes that were differential due to the miR ectopic expression, I proceeded
in a similar way that was used for the whole genome microarray analysis described in
Chapter 4. Genes were considered miR targets if they were both down-regulated upon
expression of a certain miR (miR-X) compared to Let7a and the difference of Log Fold
Change in the gene expression between ―HMF3AEcoR 38 versus 34‖ and ―Let7a versus
miR-X‖ was >1 or < -1 (equivalent to a +2 or -2 times fold change).
In theory, if the genes were up-regulated upon senescence and also upon expression of a
certain miR, this would suggest that the miR in question does not target that specific
gene and therefore down-regulation of this gene is not essential to bypass senescence.
Alternatively, if the gene is up-regulated upon senescence and down-regulated upon
miRNA expression, it could be concluded that expression of this gene is probably
important for the growth arrest.
299
Figure 6.9: Microarray profiling strategy of cells expressing miR-218, miR195, miR-193b, miR-4235p and miR-25
Triplicate independent biological samples of RNA extracted from HMF3AEcoR cells growing at 38C for 7
days and expressing the negative control Let-7a or the relevant tested micro-RNAs were analysed by
expression profiling
300
6.3.3 Microarray procedure
To perform the microarray procedure, total RNA was extracted from HMF3AEcoR cells
incubated at 38C and expressing Let7a (the reference RNA sample) or at 38C but
expressing each of the 5 miRs (miR-195, miR-423-5p, miR-25, miR-218 and miR-186).
RNA was extracted from triplicate biological cultures using Trizol (Invitrogen), frozen
and sent for analysis at the Memorial Sloan Kettering Cancer Centre/Ludwig Institute
for Cancer Research Ltd in New York.
6.3.4 MiR-186
Out of 223 predicted targets actually present in the 8064 differential data set, 75 oligos
corresponding to 51 genes (more than one oligo can represent the same gene) were
actually down-regulated by miR-186 expression more than 0.5 Log2 FC in the
microarray results and 67 of these oligos corresponding to 42 genes were also
differential (<-1 or >1, see above) between ―HMF3AEcoR 38 versus 34‖ and ―Let7a
versus miR-X‖.
Interestingly, a lot more genes were down regulated by the expression of miR-186 than
the predictive target list. These could very well be secondary targets of miR-186. Among
these are: BCL2L1, BTG2, GRAMD3, IL32, LTBP3, SGTB all of which have shown to
rescue the cells when silenced. MiR-186 could very well be rescuing by silencing these
genes.
6.3.5 MiR-195
Out of 631 predictive targets present in the 8064 differential data set, 159 oligos
corresponding to 121 genes were actually down-regulated more than 0.5 Log2 FC in the
microarray results.
301
This list includes the genes BTG2, GRAMD3, LUZP1 and AK3L1, three of which have
shown to rescue the cells when silenced with lenti-shmiRs. MiR-195 rescue could be
partly due to the silencing of these genes.
Some genes were down regulated by the expression of miR-195 but are not in the
predictive target list. Among these were CLCA2, IL32, LTBP3, SGTB and TXNIP, all
of which have shown to rescue the cells when silenced. Rescue by miR-195 rescue could
also very well be due to the silencing of these genes.
6.3.6 MiR-25
Out of 631 predictive targets present in the 8064 differential data set, 143 oligos
corresponding to 98 genes were actually down-regulated more than a 0.5 Log2 FC in the
microarray results.
This list includes the genes BTG2, GRAMD3, LUZP1 and AK3L1, three of which have
shown to rescue the cells when silenced with lenti-shmiRs.
Some genes were down regulated by the expression of miR-25 but do not appear in the
predictive target list. Among these were RUNX1, BCL2L1, ATXN10, AK3L1, LTBP3,
LUZP1, IL1 A and B, IL32, SCN2A, TXNIP.
6.3.7 MiR-218
Out of 476 predictive targets present in the 8064 differential data set, 142 oligos
corresponding to 106 genes were actually down-regulated more than a 0.5 Log2 FC. This
includes SCN2A and AK3L1 which have shown to rescue the cells when silenced with
lenti-shmiRs. MiR-218 rescue could be partly due to the silencing of these genes. Some
genes were down regulated by the expression of miR-218 but are not in the predictive
302
target list. Among these were LUZP1, LTBP3, SGTB and CLCA2, GRAMD3 most of
which were shown to rescue the cells when silenced.
6.3.8 MiR-423-5p
Out of 59 predictive targets present in the 8064 differential data set, 13 oligos
corresponding to 10 genes were actually down-regulated more than a 0.5 Log2 FC. Two
genes from this list, namely MAP1LC3A and MDX4, had their expression affected by
all constructs that were able to bypass senescence except E1A. It is however important
to note that this micro-RNA doesn‘t have as many predicted targets as for the others
micro-RNAs studied here explaining the small gene-list that overlap with the prediction.
6.3.9 MiR-372
Out of 261 predictive targets present in the 8064 differential data set, 59 oligos
corresponding to 58 genes were actually down-regulated more than a 0.5 Log2 FC.
Interestingly, TXNIP and RUNX1 belong to this list. IL1 A and B, IL6, IL32, BCL2L1
and BTG2 were down-regulated by ectopic expression of miR-372 but do not appear on
the predicted list.337 genes are down-regulated by at least - 0.5 log2 FC by all the 6
miRs including ADAMSTL1, BLNK, CLCA2, IKBKB, JAK1, MDM2, RUNX1 and
SCN2A.
6.4 RAS INDUCED PREMATURE SENESCENCE
6.4.1 Objectives
Agami and colleagues have shown that expression of miR-372/373 in conjunction with
hTERT overcomes RAS induced premature senescence (Voorhoeve, le Sage et al.
303
2006). The objectives were to test the 5 miRs plus miR-372 whose expression permitted
the bypass of the growth arrest to see if their expression in primary fibroblasts could
overcome RAS induced premature senescence.
6.4.2
Strategy
5 down-regulated miRNAs plus miR-372 into primary BJ human fibroblasts in
conjunction with hTERT by amphotropic retroviral infection. Cells were first selected
for hygromycinB to select for transduction with hTERT and then selected with
Blasticidin to isolate cells transduced with the miRNA. The cells were then challenged
with inducible ER-RAS as it has been shown that expression of an activated oncogene
such as RAS or RAF results in premature senescence in normal cells but transformation
in most immortal cells (Elenbaas, Spirio et al. 2001; Serrano and Blasco 2001; Campisi
2005; Campisi and d'Adda di Fagagna 2007). Cells derived with hTERT alone undergo
premature senescence upon challenge with these oncogenes (Morales, Holt et al. 1999)
even if they are immortal as is the case in BJ cells.
This was the exact strategy used by Agami and colleagues to demonstrate that miR372/373 immortalise human cells in conjunction with hTERT. Additional positive
controls were also tested in a similar manner: WT SV40 LT, p21 shRNA, p53 shRNA,
miR-372 as well as negative controls: hTERT + RAS alone or even RAS alone. Since it
has been suggested that primary human fibroblast are immortalised more efficiently
under low oxygen, these experiments were carried out under normal oxygen conditions
to maintain stringency. An inducible version of RAS, ER-RAS (kindly provided by Dr.
Jesus Gil), in which RAS expression can be turned on by adding 200nM 4OHT to the
growth medium was used.
304
6.4.3 Procedure
Viral supernatants were prepared using Phoenix Ampho cells for each of the following
constructs: miR-25, miR-372, miR-218, miR-193b, miR-423-5p, miR-195, WT SV40
LT, p53 shRNA, p21 shRNA. 20 µg of each of the miRs constructs were packaged
whereas 10µg of the other constructs were packaged. 10µg of the ER-RAS construct
was also packaged in Phoenix Ampho cells. hTERT viral supernatant was prepared from
a stable TEFLYA cell line that produces hTERT virus (O'Hare, Bond et al. 2001).
BJ cells were infected with each of the above described constructs and hTERT as
described in Table 6.6: 10ml viral supernatant for hTERT and 40ml of miR viral
supernatant or 10ml for LT, p53 shRNA and p21 shRNA. 40ml of miR viral supernatant
was used because it had been observed that the miR vectors produced low amounts of
viruses. Each infection was performed as biological triplicates. The cells were then
selected with hygromycin at 50 μg/ml (for hTERT alone), for at least 10 days, and then
with blastocidin at 2.5 µg/ml (for miR cultures), for at least 8 days; for WT LT, p53
shRNA and p21 shRNA, cultures were selected with puromycin at 1µg/ml for at least 6
days. Control non-infected BJ cells subjected to puromycin (1µg/ml), blastocidin (2.5
µg/ml) or hygromycin (50 µg/ml) died in 4, 7 and 9 days respectively.
After selection, all cultures were infected with 10 ml of ER-RAS (Table 6.6) and
selected with G418 at 0.75 mg/ ml for 10 days. Control non-infected BJ cells subjected
to 0.75 mg per ml G418 died in 9 days. Immediately after infection with the ER-RAS
virus, the cultures were transferred to phenol-red minus medium supplemented with
charcoal stripped serum. Medium lacking phenol red was used because a lipophilic
impurity contained in the phenol red has been described as a weak estrogen agonist
(Berthois, Pourreau-Schneider et al. 1986). Activation was carried out by the addition of
200nM 4OHT. Due to lack of time, the growth assays upon activation of RAS were
carried out by Ms. Katharina Wanek and will be presented in the following discussion.
305
Constructs
to test
hTERT
RAS
Number of
biological
triplicate
none
No
Yes
3
none
Yes
Yes
3
P53 RNAi
Yes
Yes
3
P21 RNAi
Yes
Yes
3
WT LT
Yes
Yes
3
miR-372
Yes
Yes
3
miR-25
Yes
Yes
3
miR-218
Yes
Yes
3
miR-195
Yes
Yes
3
miR-193b
Yes
Yes
3
miR-423-5p
Yes
Yes
3
Table 6.6: Layout of the primary BJ cells immortalization experiment
Triplicate cultures of BJ primary cells were infected with the indicated constructs, hTERT and RAS as
indicated above before being tested for immortalization by activating RAS.
306
6.5 DISCUSSION
6.5.1 Up-regulated micro-RNAs
Micro-RNAs inhibit gene expression by binding in the 3‘UTR of target mRNAs with
imperfect complementarity and preventing protein translation or promoting mRNA
degradation. Much of the progress in understanding miRNA function to date has been
from inhibition studies with antisense oligonucleotides (ASOs) or anti-miRNA
oligonucleotides (AMOs). As miRNAs are small nucleic acids, only 19–24 nucleotides
in length, ASO inhibition is considered the best and possibly the only practical approach
for specific pharmacological inhibition of their function (Esau 2008). ASOs targeting
mRNAs have been widely used to evaluate gene function in vitro and in vivo
(Stepkowski, Qu et al. 2000; Zellweger, Miyake et al. 2001; Watts, Manchem et al.
2005; Lee, Dunham et al. 2006) and several antisense therapeutics are currently
undergoing clinical trials .
More recently, modified AMOs were created with the dual purpose to stabilize their own
structure and to improve their affinity for their targets. Locked nucleic acid (LNA), for
example, give very strong duplex formation with their target sequences and they display
excellent mismatch discrimination, hence avoiding off-target effects (Esau 2008; Mattes,
Collison et al. 2008). A third generation of antisense oligonucleotides are
phosphodiamidate morpholino oligomers (PMO) in which the ribose ring is replaced
with a morpholine ring (Spurgers, Sharkey et al. 2008). Krützfeldt et al. linked a
cholesterol moiety to their AMOs and referred to these anti-miRNAs as antagomiRs.
AntagomiRs should be >19 nucleotides in length to provide highest efficiency in
silencing target miRNA (Krutzfeldt, Rajewsky et al. 2005). The putative therapeutic
potentials of antagomiRs were recently demonstrated in treatment of lipid metabolic
disease in animals (Esau, Davis et al. 2006). Another alternative class of AMOs is
peptide nucleic acids (PNA), which are synthetic oligonucleotides with N-(2aminoethyl)-glycine replacing ribose backbone (Fabani and Gait 2008). Finally, another
307
approach in silencing miRNA is the use of so-called micro-RNA sponge, a synthetic
mRNA that contains multiple binding sites for a particular miRNA and that is
transcribed from a plasmid containing a strong promoter (Mattes, Collison et al. 2008).
In conclusion, different classes of AMO have been shown to be efficient in silencing
miRNA and may be useful therapeutic tools.
In this case, it would have been interesting to validate the effect of the up-regulated
micro-RNAs by inhibiting those individually using antagomiRs or LNAs. They can be
introduced into cells using transfection or electroporation parameters similar to those
used for siRNAs, and enable a study of miRNA biological effects. However since they
can only be used in short term and since, the assay is a long-term one, and due to a lack
of time and an obligation to prioritise, I focussed on the down-regulated miRs, as the
effects of micro-RNAs are to silence gene expression; I expected that the downregulated miRs would target genes up-regulated upon growth arrest and also targets
identified from the shRNA screen. The 18 up-regulated miRs were still important and it
was possible to note, notably, among the up-regulated targets, 34a which has described
extensively in the last few years and been previously linked to cancer, apoptosis and
growth arrest.
6.5.1.1 MiR-34a
MiR-34a expression has been found to be reduced in human epithelial ovarian cancers
(EOC); moreover, miR-34 reconstitution in p53 mutant EOC cells resulted in reduced
proliferation, motility, and invasion (Corney, Hwang et al. 2010). These are consistent
with the data that miR-34a was up-regulated upon growth arrest. Ectopic expression of
miR-34 has also been shown to induce apoptosis, cell-cycle arrest or senescence.
MiR-34a is a known direct transcriptional target of p53, so it is not surprising that when
LT is inactivated and p53 activated upon temperature shift that miR-34a is also
increased.
308
In a recent study, the expression of miRNAs in primary human TIG3 fibroblasts after
constitutive activation of B-RAF was examined. Amongst the regulated miRNAs, both
miR-34a and miR-146a were strongly induced upon senescence indicating that miR-34a
was regulated independently of p53 during oncogene-induced senescence. Up-regulation
of miR-34a was mediated by the ETS family transcription factor, ELK1 (Christoffersen,
Shalgi et al. 2010). This totally corroborates the results of the microarray which placed
miR-146a as the top up-regulated miR and miR-34a the second. Interestingly, miR-154,
miR-376 and miR-495 are also up-regulated in both my and the Christoffersen study.
It has been shown that miR-34a regulates silent information regulator 1 (SIRT1)
expression. MiR-34a inhibits SIRT1 expression through a miR-34a-binding site within
the 3' UTR of SIRT1. This inhibition of SIRT1 leads to an increase in acetylated p53
and expression of p21 and PUMA, transcriptional targets of p53 that regulate the cell
cycle and apoptosis, respectively (Yamakuchi, Ferlito et al. 2008). This is consistent
with my finding that while miR-34a is up-regulated 3.71 log2 FC upon senescence
(Table 4.3), SIRT1 was down-regulated by 1.19 log2 FC (chapter 2).
This also in agreement with my finding that SIRT1 ectopic expression bypassed
senescence in CL3EcoR cells (See Chapter 4).
Additionally, in a recent study, cellular senescence was shown to be induced by nutlin3a, an MDM2 inhibitor, in normal human fibroblasts. Nutlin-3a acts by up-regulating the
expression of miR-34a, miR-34b, and miR-34c through the activation of p53 and the
repression of ING2 (inhibitor of growth 2) (Kumamoto, Spillare et al. 2008).
6.5.1.2 MiR-146a
The most highly up-regulated miR is miR-146a with a Log2 fold change of 4.33 which is
more than 20 times the expression it has in growing cells. Recently, miR-146a
expression has been shown to be up-regulated by IL-1β (Perry, Williams et al. 2009)
which is one of the top up-regulated genes upon senescence (Chapter 4, Table 4.1B).
MiR-146 has also been reported to be up-regulated by Breast cancer metastasis
309
suppressor 1 (BRMS1), a predominantly nuclear protein that differentially regulates
expression of multiple genes, leading to growth arrest and suppression of metastasis
(Hurst, Xie et al. 2009). The link between miR-146a up-regulation and growth arrest in
the Hurst study is in agreement with my results. In addition, miR-146 was also found to
be an NF-κB dependant gene (Taganov, Boldin et al. 2006) which is consistent with the
activation of NF-κB signalling having a causative role in promoting cellular senescence.
Interestingly, Bhaumik et al (Bhaumik, Scott et al. 2008) have suggested that miR146a/b can act as a negative regulator of NF-κB activity in Breast Cancer cells. Taken
together, these results suggest that there may be an autoregulation loop where NF-κB
activates miR-146 which then suppresses NFκB activity.
6.5.2 Down-regulated micro-RNAs
Expression profiling indentified 15 micro-RNAs that were down-regulated upon
senescence (Table 6.5). Many of them were up-regulated upon quiescence. These downregulated micro-RNAs and their importance in the senescence process were, when
available for ectopic expression, functionally analysed in the HMF3A cell system by
constitutive expression and followed by growth complementation assay.
12 of these miRs plus an extra 6 miRs were chosen for ectopic expression in a
complementation assay and some of them yielded large numbers of growing colonies by
promoting cell growth namely miR-195, miR-25, miR-193b, miR-186, miR-218 and
miR-423-5p.
6.5.2.1 MiR-25
Replicative senescence was shown in vitro to be associated with the decrease of miR15b and miR-25 among others in an HDF model. This decrease was shown to elicit the
increase of MKK4 a pivotal upstream activator of c-Jun N-terminal kinase and p38
which are essential to the induction of cellular senescence (Marasa, Srikantan et al.
2009). This data is in accordance with my results that miR-25 is down-regulated upon
310
cellular senescence. Loss of miR-25 expression in the HMF3A cells could result in an
increase of MKK4 leading to induction of senescence and thus ectopic expression of
miR-25 would lead to down-regulation of MKK4. However, MKK4 (or MAP2K4) is not
one of the predicted targets of miR-25 and is not present amongst the genes found to be
up-regulated suggesting that in the HMF3A cells, modulation of MKK4 expression is
unlikely to be the mechanism by which senescence is triggered.
Furthermore, miR-25 has also been reported down-regulated in ASM (Airway smooth
muscle) cells exposed to IL-1β (Kuhn, Schlauch et al. 2010). This is consistent with my
finding that IL-1β was one of the top up-regulated genes upon senescence and miR-25
was also down-regulated (Chapter 4, Table 4.1B) (Table 6.5). MiR-25 may be downregulated by the increase in IL-1β. But, this would imply that miR-25 is downstream of
the activation of the NF-κB pathway. I have shown here that miR-25 down-regulation
was causal to senescence because it can be bypassed upon ectopic expression of miR-25.
In addition, IL-1α and IL-1β are both down-regulated by the expression of miR-25.
Therefore, miR-25 cannot be considered as merely a consequence of the regulation of
IL-1β upon senescence but as part of a pathway, maybe a negative loop of modulation
necessary to activate the growth arrest.
The miR-106b-25 polycistron, which is located within the 3‘UTR of the MCM7
transcript, was recently reported to exert potential proliferative, anti-apoptotic, cell
cycle-promoting effects in vitro and tumourigenic activity in vivo. In this study, miRs-93
and -106b targeted and inhibited p21CIP1/WAF1/Sdi1, whereas miR-25 targeted and inhibited
the pro-apoptotic factor Bim.
This polycistron was shown to be up-regulated progressively at successive stages of
neoplasia, in association with genomic amplification and over-expression of MCM7
(Kan, Sato et al. 2009). This is accordance to my finding that miR-25 is down-regulated
upon senescence (Table 6.5) and that its ectopic expression promotes cell growth (Figure
6.4 A and B, 6.5, 6.6B, 6.7C and 6.8B). In addition, MCM7 is also down-regulated upon
311
senescence by 1.4 log2 FC in the HMF3AEcoR cells and becomes up-regulated when
senescence is overcome probably resulting in an increase in miR-25 expression.
6.5.2.2 MiR-195
MiR-195 expression has been linked in several studies to Cancer and tumourigenesis. In
a recent study, cancer-specific miRNAs significantly altered in the circulation of breast
cancer patients were detected and increased systemic miR-195 levels in breast cancer
patients were reflected in breast tumours. Furthermore, circulating levels of miR-195
and Let7a were shown to decrease in cancer patients postoperatively, to levels
comparable with control subjects, following curative tumour resection (Heneghan,
Miller et al. 2010) suggesting that expression of miR-195 is strongly involved in the
development and maintenance of Breast Cancers. This is in accordance with my findings
that miR-195 is down-regulated upon senescence (Table 6.5) and that its ectopic
expression promotes cell growth (Figure 6.4B, 6.5, 6.7A and B and 6.8 A and B).
Another study recently found that perturbation of the miRNA pathway function in
human embryonic stem cells (hESCs) by RNA interference-mediated suppression of
DICER and DROSHA, 2 proteins essential in the biogenesis of all miRs, attenuates cell
proliferation. In this study, normal cell growth can be partially restored by introduction
of the mature miR-195 and miR-372 which regulate two tumour suppressor genes:
WEE1, a negative G2/M kinase modulator of the CycB/CDK complex and CDKN1A,
which encodes p21CIP1/WAF1/Sdi1, the cyclin dependent kinase inhibitor. WEE 1 levels
control the rate of hESC division, whereas p21CIP1/WAF1/Sdi1 levels must be maintained at
a low level for hESC division to proceed (Qi, Yu et al. 2009). These data support the
result that introduction of miR-195 is sufficient to bypass senescence in HMF3A cells
(Figure 6.4B, 6.5, 6.7A and B and 6.8 A and B). However, these results are
contradictory to my results indicating that WEE1 was actually down-regulated upon
senescence and gets up-regulated when senescence is overcome including with the
ectopic expression of miR-195. This suggests that miR-195 may not directly target
WEE1 and utilise a different pathway.
312
6.5.2.3 MiR-218
In a recent study, miR-218 expression was shown to be reduced significantly in gastric
cancer tissues, in H. pylori-infected gastric mucosa, and in H. pylori-infected AGS cells.
In the same study, over-expression of miR-218 inhibited cell proliferation and increased
apoptosis in vitro (Gao, Zhang et al. 2010). Epidermal growth factor receptor coamplified and over-expressed protein (ECOP), which regulates NF-κB transcriptional
activity and is associated with apoptotic response. NF-κB transcriptional activation and
the transcription of ECOP and cyclo-oxygenase-2 (COX2), a proliferative gene
regulated by NF-κB, were all described to be targets of miR-218 (Gao, Zhang et al.
2010). This suggests that upon expression of miR-218, the NF-κB pathway, ECOP and
COX2 expression should be down-regulated. In addition, COX2 expression had been
previously described as linked with an increased risk of bladder cancer and prostate
cancer (Kang, Kim et al. 2005) and its inhibition was shown to promote growth arrest in
colon cancer cells and prostate cancer cells (Grosch, Tegeder et al. 2001; Narayanan,
Narayanan et al. 2006). It would be, therefore, logical for COX2 to be down-regulated
upon growth arrest.
MiR-218 was also shown to be involved in cervical carcinogenesis. Its expression was
down-regulated in HPV-positive cell lines, cervical lesions and cancer tissues containing
HPV-16 DNA compared to both C-33A and the normal cervix. It was also shown that
the epithelial cell-specific marker LAMB3 is a target of miR-218 and that LAMB3
expression was increased in the presence of the HPV-16 E6 oncogene through miR-218
modulation (Martinez, Gardiner et al. 2008).
These results are completely contradictory with the data that miR-218 was downregulated by -1 log2 FC upon growth arrest (Table 6.5) and that its ectopic expression
promoted cell growth (Table 4.6A, B and C). In addition, expression of miR-218 does
not seem, in this case, to inhibit the NF-κB pathway as it is described in the Gao paper,
as downstream targets of NF-κB such as IL-1A, IL1-B, IL-8, BMP2 and SOD2 are upregulated in cells expressing miR-218 as shown in the microarray analysis. NF-κB
pathway was activated upon growth arrest but not reversed by miR-218 expression and
313
neither ECOP nor COX2 (or PTGS2) nor LAMB3 expression varied upon growth arrest.
This suggested that modulation of ECOP, COX2 and LAMB3 through miR-218
modulation was unlikely to be the mechanism by which senescence was triggered and
that modulation of NF-κB was not essential for abrogation of the growth arrest through
miR-218 expression.
6.5.2.4 MiR-193b
The literature is not clear with respect to the regulation of miR-193b expression. In one
study, miR-193b was proposed to be up-regulated in Hepato-cellular cancer (HCC). In
this study, HepG2 malignant hepatocytes were stably transfected with full-length HCV
genome (Hep-394) or an empty vector (Hep-SWX) and micro-RNA expression profiling
performed on both cell types. MiR-193b was shown to be over-expressed 5-fold in Hep394 cells compared to the control and to target Mcl-1, an anti-apoptotic protein (Braconi,
Huang et al. 2010).
In three other studies, however, miR-193b was reported to be down-regulated in cancer.
One study looked at the miRNA expression profile in 10 pairs of endometrioid
adenocarcinoma and adjacent non tumourous endometrium and found that miR-205,
miR-449, and miR-429 were greatly enriched whereas miR-204, miR-99b, and miR193b were greatly down-regulated in adenocarcinoma tissues (Wu, Lin et al. 2009). In
another study, miR-193b was also down-regulated in clinical prostate cancer samples
compared to benign prostatic hyperplasia. In this study, in addition, it was shown that
expressing miR-193b in 22Rv1 cells using pre-miR-193b oligonucleotides caused a
significant growth reduction (p<0.001) resulting from a decrease of cells in S-phase of
the cell cycle (p<0.01). The authors even proposed that miR-193b could be an
epigenetically silenced putative tumour suppressor in prostate cancer (Rauhala, Jalava et
al. 2010). Expression of miR-193b was also shown to be down-regulated during breast
cancer cell metastasis (Li, Yan et al. 2009). It is possible that the observed differences
are due to the type of cancer, cell line/tissue and the context.
314
In our HMF3A cells, miR-193b shows a down-regulation upon senescence (Table 6.5)
and its ectopic expression promotes cell growth and bypasses senescence (Figure 6.5 and
6.8A and B).
Since senescence bypass with miR-193b was weaker than with the other miRs, cells
expressing miR-193b were not profiled by microarray analysis, so it was not determined
what changes in expression were caused by miR-193b expression.
MCL-1, which was a proposed direct target of miR-193b in the Rauhala study, was not
modulated upon growth arrest which suggests that its modulation is not the cause of
neither the abrogation nor the trigger of growth arrest.
6.5.2.5 MiR-186
MiR-186 has been found to be significantly up-regulated in most pancreatic cancer
tissues and cell lines, in conjunction with seven others miRs: miR-196a, miR-190, miR186, miR-221, miR-222, miR-200b, miR-15b, and miR-95 (Zhang, Li et al. 2009).
Levels of miR-186 and miR-150 were also reported to be higher in cancer epithelial
cells than in normal cells. This study also showed that increased expression of miR-186
and miR-150 in cancer epithelial cells decreases P2X7 mRNA by activation of miR-186
and miR-150 instability sites located at the 3'-UTR-P2X7. Indeed, treatment with
inhibitors of miR-186 and miR-150 increased P2X7 mRNA level (Zhou, Qi et al. 2008).
These results are in accordance to my findings that miR-186 expression is downregulated upon senescence (Table 6.5) and that its ectopic expression bypasses the
growth arrest (Table 6.5 and 6.8A and B). The levels of P2X7, however, do not vary
upon senescence or upon rescue, which suggest that its expression does not play a major
role in the senescence pathways.
6.5.2.6 MiR-423-5p
A recent study investigated the significance of miRNAS in patients with locally
advanced head and neck squamous cell carcinoma and found miR-423, miR-106b, miR20a, and miR-16 to be up-regulated and miR-10A to be down-regulated.(Hui,
315
Lenarduzzi et al. 2010) This is in accordance with my finding that miR-423 is downregulated upon senescence (Table 6.5).
However, another study carried out to analyze the miRNA expression profile of 17
malignant mesothelioma samples using miRNA microarray has obtained contradictory
results. Malignant mesothelioma (MM) is an aggressive cancer arising from mesothelial
cells, mainly due to asbestos exposure. MiR-423 was found to be down-regulated in
tumour samples compared with normal sample along with miR-34a (Guled, Lahti et al.
2009).
Although the miR-34a results are in accordance with my data, the miR-423-5p results
are contradictory. However, after careful reading of the Guled paper, it was noted that at
two different places miR-423 was replaced by miR-429 and thus it remains to be
verified exactly which miRNA is down-regulated.
6.5.3 Expression profiling of HMF3A cells in which senescence has been
bypassed by ectopic expression of miRs
An interesting fact is that when looking at the top up-regulated genes upon senescence,
mostly, the genes were down-regulated upon rescue by abrogation of the pRb pathway,
the p53 pathway or expression of miRs. However, some genes stood out because they
were also up-regulated by the expression of 3 to 4 miRs namely miR-186, miR-195,
miR-218 and miR-423. Among these are IL-1A, IL-1B (with 2 different oligos), BMP2,
CEBPD, SOD2, IL-15RA, IL-6, CCL20, CSF2, RAB27B, BIRC3, DUSP6 all of which
are either direct targets of NF-κB or linked to the activation of NF-κB. This suggests that
the down-regulation of NF-κB may not be necessary in the rescue with the miRs miR186, miR-195, miR-218 or miR-423. Therefore maybe these four micro-RNAs use a
different pathway to bypass senescence than miR-25 or miR-372.
It is also interesting to note from the results of the microarrays with all miRs, that the
miR-372 expressing cells expression pattern was very different to that of miR-186, miR316
25, miR-195, miR-423-5p and miR-218. For example, when looking at the list of the top
genes down-regulated by miR-195 expression, it was possible to see a large majority of
the targets expression were regulated identically by the others 4 miRs mentioned by
were either up-regulated or not regulated by miR-372. It is also possible to note that
among these first 100 targets several were actually NF-κB targets such as SOD2, SFS2,
IL1A and B, CCL20, CEBPD, IL6 and IL8.
When overlapping all the miRs results, it was observed that ADAMTSL1 is a confirmed
target of 4 miRs out of 6 and is down regulated by all 6 miRs. There are actually 337
genes that are down-regulated by all 6 miRs including CLCA2, IKBKB and MDM2.
6.5.4 Expression of the miRs in 226L cells
226L8/13 cells correspond to human breast epithelial cells which have been
immortalized by introducing into them the U19tsA58 LT and hTERT that were used to
derive the HMF3A cells.
In an analogous manner to the HMF3A cells, these cells are immortal at 34°C but stop
dividing upon inactivation of SV40 LT antigen at 38°C. These cells are used by
Katharina Wanek for her thesis and she has prepared clones of these cells by stably
infecting them with the murine ecotropic receptor and identified clone #7 cells as clonal
cell line for identifying senescence pathways. These cells responded to abrogation of the
p53 or pRb pathway in a similar manner to the one of the HMF3A, by bypassing the
growth arrest.
After my results of the miRs expression with the HMF3A cells, it was interesting to test
the same constructs in the mammary epithelial cell line to see if their action was only
limited to fibroblasts. The 226L cells were transduced with the miRs-vectors miR-25,
miR-218, miR-193b, miR-423-5p, miR-186, miR-186 as well as miR-372 and miR-373
as positive controls and miR-15a and Let7a as negative controls respectively. The cells
317
were selected for expression, then shifted to the non permissive temperature for 3 weeks
and stained as for the HMF3A protocol.
The results obtained by Katharina Wanek (Figure 6.10) show similar results for all of the
miRs tested apart from miR-193b and miR-186 which did not display any rescue. The
positive controls miR-372 and miR-373 expression show very good cell growth with
many growing (blue coloured) colonies and there is very little background in the
negative controls, Let7a and miR-15a. MiR-373, surprisingly, works better than miR372 in the 226L epithelial cells which is the opposite in the HMF3A fibroblasts. This
variation could be due to the cell type.
The most efficient miR expression here seems to be miR-195 which corroborates my
own results followed by miR-25, miR-218 and miR-423-5p.
MiR-195 show less
colonies than miR-373 but they are much darker and larger probably suggesting that
they are growing faster. The reduced number of colonies is, however, not due to reduced
infection since stably transduced cells were reseeded.
These results reinforce the idea that these 4 miRs are important effectors with a direct
causality in the gene regulation changes that happen during senescence. It also means
that miRs involvement in senescence is not limited to a certain cell type. There may be
slight differences between epithelial cells and fibroblasts but the same sorts of miRs
seem to be functional.
6.5.5 RAS transformation of primary cells
This experiment is discussed here as it took months to optimise and for timescale
reasons, the final growth assays were performed by Katharina Wanek.
318
Figure 6.10: Ectopic expression of micro-RNAs in human breast epithelial cells
226L cells were infected in triplicate with retrovirus expressing the indicated miRs expression constructs
and assayed for growth complementation at 38°C in T-75cm2 flasks. After 3 weeks the flasks were
stained and photographed.
319
In this experiment, miR-Vectors capable of ectopically expressing each of the 6 downregulated miRNAs were introduced by retroviral delivery (Ampho) into primary BJ
human fibroblasts in conjunction with hTERT. Cells were selected hTERT and the
miRNAs before being transduced with inducible ER-RAS. Positive controls included
WT LT, p21 shRNA, p53 shRNA and miR-372 and negative controls included hTERT +
RAS alone or even RAS alone. The cells were then grown for 7 days in medium
containing 200nM 4OHT as in to activate the expression of RAS. Then the cells were
reseeded in 96 well plates in triplicate for the growth assay. The cells were counted at
day 1, day 5 and day 7.
As the infection were already done in biological triplicate and the cell count was done in
triplicate for each cell flask condition, there was 9 results per condition in total and the
data looked surprisingly tight for a tissue culture results.
At day 5, it was already possible to see a large difference between p53 shRNA, WT LT
and miR-372 positive control and the rest of the cells with about 70% more cells in
average in these three cultures than the rest.
At day 7, the difference was even more pronounced. The curve with RAS alone
remained flat and the cells became growth arrested and displayed a senescence like
phenotype. This is in accordance with the Gil lab that kindly provided us with the ERRAS construct that RAS expression causes premature senescence in BJ cells (Barradas,
Anderton et al. 2009).
RAS + hTERT expressing cells registered a big decrease in their growth rate and looked
arrested, however the numbers were higher than with RAS alone, suggesting that
hTERT provides a boost to the cell growth. This is not surprising since hTERT
immortalizes BJ cells.
Surprisingly, p21 shRNA expressing cells did not seem to grow at a better rate than
hTERT + RAS expressing cells even though it is one of the best candidates for
320
bypassing senescence in the HMF3A cells. The p53 shRNA derived cells were resistant
to the RAS effect and continued to divide.
In human, some studies have shown that inactivation of p21CIP1/WAF1/Sdi1 alone was
sufficient to bypass OIS (oncogene-induced senescence). A recent study showed that
siRNA-mediated knockdown of p21CIP1/WAF1/Sdi1 rescues from Ras-induced senescence
in human mammary epithelial cells (HMECs) (Borgdorff, Lleonart et al. 2010). This is
in contradiction with my findings. In addition, inhibition of p21CIP1/WAF1/Sdi1 expression
in BJ foreskin human fibroblasts also resulted in RasG12V-resistant growth (Voorhoeve,
le Sage et al. 2006). However, another study showed that, like in our case, inactivation
of p21CIP1/WAF1/Sdi1 alone in LF1 human lung fibroblasts could not bypass RasG12Vinduced senescence unless p16INK4a was inactivated as well (Wei, Herbig et al. 2003).
In mice, my results are in accordance with the findings of the Serrano lab (Pantoja and
Serrano 1999) that p21-deficient murine fibroblasts are not efficiently transformed by
oncogenic Ras, and this is in contrast to p53-/- equivalent cells that are efficiently
transformed, indicating that p21CIP1/WAF1/Sdi1 is not essential for the anti-proliferative
response induced by moderate levels of oncogenic Ras, and that p21-deficient
fibroblasts are refractory to transformation. Regarding other cell types different than
fibroblasts, it should be mentioned that there are conflictive reports about the
susceptibility of p21-deficient keratinocytes to be transformed by oncogenic Ras
(Missero, Di Cunto et al. 1996; Weinberg and Yuspa 1997). Missero reports that
primary keratinocytes derived from p21CIP1/WAF1/Sdi1 knockout mice, can be transformed
with a ras oncogene while Weinberg reports the contrary. The mice were from different
strains and I can hypothesize that the genetic context might be of importance. It can be
concluded from this discordant results that p21CIP1/WAF1/Sdi1 silencing is not always
sufficient along with hTERT expression to overcome RAS induced premature
senescence and that it is likely to depend upon the genetic context, the cell type, and the
organism type.
321
It was surprising that p21CIP1/WAF1/Sdi1-/- cells did not transform with RAS like in the
Voorhoeve paper but underwent growth arrest. The p53 silenced cells, however, did
transform with RAS indicating that even though in terms of abrogating growth arrest in
the conditional immortal cells, p21CIP1/WAF1/Sdi1-/- cells behaved the same as p53-/-, in
case of the primary cells, they are not the same. This suggests that p53 has others targets
in addition to p21CIP1/WAF1/Sdi1.
Ectopic expression of miR-372 overcame Ras induced senescence in conjunction with
hTERT and resulted in the cells being transformed. This is in accordance with the data
published by Voorhoeve et al.
Preliminary evidence suggests that only the miR-423-5p expressing cells continued
dividing at a growth rate above hTERT reconstituted BJ cells. The remainder stopped
dividing. Further experiments are now underway to confirm this finding.
Although miR-423-5p was the only micro-RNA able to overcome RAS induced
premature senescence in BJ cells, it does not mean that it is more important than the
other miRs in the senescence process but it follows a different pathway than the others, a
pathway shared with p53 or LT. Indeed, p21CIP1/WAF1/Sdi1 which is a very important
effector in the senescence process could not overcome Ras induced senescence in these
cells either (Brown, Wei et al. 1997; Wei and Sedivy 1999).
6.5.6 Further work
An important question would be to determine what is causing the change in expression
of the miRNAs. The regulation could be at the level of transcription or at the level of
processing. For example, miR-25 belongs to the miR-106b-25 cluster of three miRNAs
derived from the MCM7 transcript (Tanzer and Stadler 2004)( www.miRbase.com);
MCM7 is down-regulated >2 fold upon HMF3A growth arrest. In our HMF3A cells,
miR-25 was down-regulated >2 fold whereas miR-106b and 93 were not significantly
322
down-regulated. Their processing might therefore be regulated differently. The
regulation could also be at the level of transcription since miRNAs are transcribed by
RNA Polymerase II and thus it will be important to identify the transcription factors
involved. For example, Mcm7 expression has been proposed to be E2F dependent
(Suzuki, Okuyama et al. 1998; Bruemmer, Yin et al. 2003) but I have also found it be
modulated by p53 as well as p21 CIP1/WAF1/Sdi1 in the HMF3A cells upon growth arrest.
So what regulates MCM7 expression in cellular senescence?
However, the main question would be to focus on the miR-423-5p. It was differential in
the HMF3A upon growth arrest; its ectopic expression bypassed growth arrest. It also
overrides Ras induced premature senescence in primary BJ cells. The next step could be
to try to dissect the details of what makes the difference compared to the other miRs in
the transformation of primary fibroblasts. It would be interesting as well to look at the
gene expression and protein levels of members of the senescence pathway such as p53,
p21CIP1/WAF1/Sdi1, pRb but also effectors of the NF-κB pathway in cells expressing miR423-5p.
323
7
SUMMARY AND FINAL DISCUSSION
Cellular senescence is an irreversible program of cell cycle arrest that normal cells
undergo both in vitro and in vivo in response to a variety of intrinsic and extrinsic
stimuli. Senescence is associated with organismal ageing as it promotes the disruption
of tissue renewal and repair processes as well as the depletion of progenitor cell
populations. Senescence also represents an important tumour suppressive mechanism
that limits the growth capacity of potentially cancerous cells. Bypass of senescence
therefore represents a mechanism by which cells can overcome finite proliferative
potential, one of the six proposed hallmarks of cancer cells (Hanahan and Weinberg
2000).
In conjunction with hTERT, LT can immortalise many human cells including primary
human fibroblasts by inactivating the p53-p21 and p16-pRB tumour suppressor
pathways. Consequently, the generation of a thermolabile mutant of LT, U19tsA58, led
to the development of a conditionally immortalised human mammary fibroblast cell
model, HMF3A (O'Hare, Bond et al. 2001). HMF3A cells and its clonal derivative,
CL3EcoR cells, grow at 34C, but undergo an irreversible growth arrest within 5 days
upon temperature shift to 38C. Since telomerase remains constitutively active in these
cells at both 34C and 38C, the growth of HMF3A cells is entirely dependent upon LT
activity suppressing p53 and pRb activities.
Previously, Hardy et al (2005) have carried out a 6000 genes microarray analysis to
identify changes that occur upon induction of irreversible growth arrest in the HMF3A
cells and found that some of the changes in expression directly correlated with the
transcriptional changes that are induced upon replicative senescence in normal human
mammary fibroblasts.
Moreover, RNA interference and in silico analysis had indicated an important role for
the p53, pRb and NF-κB signalling pathways in this process.
324
7.1 SUMMARY OF RESULTS
The main goal of this thesis was to profile the expression changes upon senescence in
order to find new targets which would help to dissect the senescence pathways and then
to analyse them in detail.
Since the cells growth-arrest in a synchronous manner, I have used Affymetrix
expression profiling to identify the genes differentially expressed upon senescence. This
identified 816 up- and 961 down-regulated genes whose expression was reversed when
growth arrest was abrogated. Overlay of this data set with the meta-signatures of genes
up-regulated in cancer showed that 50% of them were down-regulated upon senescence.
Remarkably, 65 of the up- and 26 of the down-regulated genes are known downstream
targets of NF-κB indicating that senescence may be associated with activation of the
NF-κB pathway. Perturbation of this pathway by direct silencing of NF-κB subunits, by
positive and negative upstream modulation, or by expressing the super repressor of NFκB, can overcome growth arrest indicating that activation of NF-κB signalling has a
causal role in promoting senescence.
Moreover, this activation of NF-κB upon senescence could also be the cause of the
down-regulation of FOXM1 and E2F and consequently of their downstream targets that
are critical for cell cycle progression particularly in the G2 phase.
At the same time, I also applied a retroviral shRNA screen covering ~10,000 genes to
the same CL3EcoR cell model. Overlapping these results with the microarray data
revealed particularly interesting targets, such as LTBP3, LAYN, SGTB, TMEM9B and
ATXN10 which were both up-regulated upon senescence and able to bypass growth
arrest when silenced by shRNA expression.
I also profiled micro-RNA expression. 15 of the top micro-RNAs down-regulated upon
senescence were chosen for ectopic expression in the HMF3A cells. MiR-25, miR-4235p, miR-218, miR-186, miR-193b and miR-195 upon ectopic expression were able to
325
bypass the growth arrest. Subsequently, these micro-RNAs were introduced into BJ
human primary fibroblasts along with hTERT and activated RAS and miR-423-5p only
was found to bypass RAS induced senescence.
In conclusion, my work has uncovered novel markers involved in senescence as well as
identifying that both activation of p53-p21 and p16-pRB pathways results in activation
of NF-κB signalling which promotes senescence. Both results lead to a better
understanding of senescence and the underlying signalling pathways.
7.2
FUTURE DIRECTIONS
Due to the time constraints and multiple aspects of the project, not all candidate genes
identified by the different experimental approaches were functionally validated. Instead,
a prioritisation of the identified targets was made and only a few were chosen for further
investigation. For each of these targets, ectopic expression or silencing by RNAi was
employed to complement the conditional growth of HMF3A EcoR or CL3EcoR cells. Upon
confirmation of this activity, experimental analysis should be extended to primary
human fibroblasts and other primary human cells, similarly to the experiment with the
micro-RNAs. Since activities of each these genes may be impaired in different tumour
types, expression could also be analysed in a variety of primary human tumours and
cancer cell lines.
7.2.1 Saturation of shRNA screen in CL3EcoR
The RNAi screen (see Chapter 3) should be performed on a much larger scale, as there
was substantial evidence to suggest that the screen was performed under non-saturating
conditions. Using a higher volume of virus and a higher concentration of antibiotics for
selection could also ensure higher expression level of the inserts and perhaps the
expression of shRNA that were not sufficiently expressed in our screen.
326
Although improvements in shRNA technology will be required to perform fully saturating loss-of-function screen, the development of multiple fully validated shRNA
libraries coupled with the interrogation of a larger number of cell lines would permit
saturating genetic screens.
7.2.2 Secondary shRNA screen
The primary shRNA screen uncovered 111 different genes and another 30 inserts
corresponding to unidentified loci. Most of these targets were not investigated since they
did not overlap with the microarray data. However, it is always possible that some
targets could be activated without having their mRNA levels affected for example by
loss of an upstream inhibitor.
To address this issue, a secondary screen should be performed (as suggested in Chapter
3) with the Lentiviral shMirs library on these 111 genes to functionally validate the
primary screen and eliminate false positives. This would identify any genes whose
expression or activation was causal to senescence.
7.2.3 Ectopic expression validation by protein analysis
Ectopic expression of genes found to be down-regulated upon senescence was not
verified for all constructs and therefore it was not possible to conclude whether
expression of HMGB2, DEPDC1, MELK and NEK2 would bypass senescence. It would
be very valuable to find antibodies that can specifically identify protein expression for
these genes and determine whether differential microarray expression correlates with
differential expression at the protein level.
Similarly, for the shRNA silencing, it would be good practice to verify that silencing
induces changes in the protein expression.
327
Alternatively, if these constructs do not give a satisfactory level of expression, other
constructs could be designed and prepared for use in the CL3EcoR cells.
Despite any protein expression evidence, the possibility that LT-mediated negative
regulation of cellular proteins may have occurred in the HMF3A system cannot be
discounted. This is a particularly important point to consider when determining the
effectiveness of expression knockdown by shRNAi; down-regulation of expression by
shRNA was performed for TRIB2, GRAMD3, CDKN2A, RUNX1, BLCAP, SCN2A,
CLCA2 and AK3L1 but the corresponding protein levels have not been assessed. Also,
whilst the results of the shRNA screen were validated for 5 targets by using multiple
shRNA constructs for each gene, again protein levels were not verified. By extension
therefore, it is possible that stable proteins may not be identified by functional screens,
such as RNAi screens. Therefore, it may be important to consider the utilisation of
alternative strategies to specifically target protein activity such as short peptide
inhibitors or dominant-negative peptides (for example, GSEs).
7.2.4 FOXM1
FOXM1 was one of the most highly down-regulated gene upon senescence in the
microarray data. This was also supported by our finding that constitutively active
FOXM1 abrogated senescence in the CL3EcoR cells. Moreover, in another study, acute
activation of NF-κB was shown to trigger growth arrest (Penzo, Massa et al. 2009) along
with repression of FOXM1 and genes associated with transit through G2 phase. This
makes FOXM1 a target of choice to study in understanding senescence pathways and the
involvement of NF-κB signalling.
7.2.4.1 Which spliceform is important?
At the expression level, it would be interesting to find out exactly whether the 3
FOXM1 spliceforms are all down-regulated upon senescence and whether the level of
328
down-regulation is similar for all three. The aim would be to determine if FOXM1b and
FOXM1c are both down-regulated in their expression. Since FoxM1a is potentially a
natural dominant negative, it will be important to determine if it is also differentially
expressed and how this affects the transcriptional activity of the other isoforms.
Similarly, finding which spliceform activity is required for the reversal of the growth
arrest would bring light into the mechanism of action of FOXM1 in the senescence
process. I have already shown that constitutively active FOXM1c can overcome
senescence in CL3EcoR cells but if FOXM1b is also down-regulated upon senescence, I
could attempt complementation with a constitutively active FOXM1b. Since FOXM1c
contains exon A1 which is not present in FOXM1b, it is possible they have slightly
different functional activities.
7.2.4.2 Which kinases regulate the activation of FOXM1?
I have found that senescence in HMF3A cells can be abrogated by the constitutively
active FOXM1c but not the wild type protein. This indicates a requirement for
activation of FOXM1. The N-terminus of FOXM1 contains an auto-repressor domain
that inhibits transactivation by an intramolecular interaction with the C-terminal TAD.
This repression can be relieved by phosphorylation of multiple cdk sites within the TAD
by cyclinA/cdk2 or possibly by cyclinE/cdk2; PLK1 and PLK4 may also play a role.
Our expression profiling data indicates that cdk2 and cyclinE expression are unaffected
upon growth arrest whereas cyclinA expression is down-regulated about 20 fold, PLK1
30 fold and PLK4 12 fold respectively.
To determine directly which of these kinases are important for activating FOXM1 for
abrogating senescence, co-expressing each of these kinases with full length FOXM1
could be utilized
329
7.2.4.3 What is the mechanism of action of FOXM1?
To determine the mechanism of action of FOXM1, it is critical to identify what are the
downstream targets and what is their relationship to genes found to be differentially
expressed in HMF3A upon cell senescence by expression profiling?
I have already undertaken a highly sensitive expression profiling analysis of HMF3A
cells when they undergo irreversible growth arrest. Expression profiling of cells rescued
by activated FOXM1 could be performed and overlapped with our microarray data and
the genes whose expression is maintained by expression of FOXM1 isoforms should be
targets of FOXM1.
7.2.4.4 What causes the decreased expression of FOXM1
in cell senescence?
Another important question would be to determine what causes the down-regulation of
FOXM1 upon cell senescence. Although it was recently suggested that that Stress-activated
kinase p38 (p38SAPK) is capable of inhibiting FOXM1 expression (Adam et al, 2009), the
transcription profiling data indicates that this unlikely to be the mechanism, since
expression of the three isoforms α (MAPK14), β (MAPK11) and  (MAPK13) of p38SAPK
present in HMF3A cells, is unaffected upon growth arrest.
Previously it was suggested that in Basal Cell Carcinomas, FOXM1 was a downstream
target of Gli1, which is transcriptionally up-regulated by Sonic hedgehog (Shh)-signalling
(28). Gli1 is a member of the Gli family of three transcription factors Gli1, 2 and 3. Gli 1
and 2 are activators whereas Gli3 is a repressor. The expression profiling data shows that all
three Gli proteins are expressed in proliferating HMF3A cells but upon growth arrest Gli2
and 3 are down-regulated whereas Gli1 may be slightly up-regulated. By expressing and
silencing Gli 1, 2 or 3, it would be possible to directly determine their effect on FOXM1
expression.
330
7.3 FINAL REMARKS
Cellular senescence is closely associated with cancer development.
Premature
senescence induced following activation of oncogenes or inactivation of tumor
suppressor genes (Courtois-Cox, Jones et al. 2008) is a potent anti-tumorigenic defense
mechanism. It is also known that cellular transformation by activated ras requires
cooperation from ‗immortalizing‘ oncogenes that overcome the senescence response,
such as those inactivating p53 (Land, Parada et al. 1983; Seger, Garcia-Cao et al. 2002).
Recent studies have demonstrated that senescent cells can be detected in early-stage,
premalignant lesions of lung, pancreas, skin and prostate in both human cancer patients
and mouse tumor models (Narita and Lowe 2005; Sun, Yoshizuka et al. 2007).
In addition, I have shown in this study that nearly 50% of genes that were up-regulated
in a study that analysed the meta-signatures of over-expressed genes upon neoplastic
transformation and in undifferentiated cancer (Rhodes, Yu et al. 2004) were downregulated upon senescence which further confirm the role of senescence as a barrier to
cancer development.
The fact that senescent cells have been detected in vivo provides compelling evidence
that cellular senescence represents a bona fide biological process acting as a protection
against cancer development.(Braig, Lee et al. 2005; Chen, Trotman et al. 2005; Collado,
Gil et al. 2005; Michaloglou, Vredeveld et al. 2005). This has profound implications for
the study of both organismal ageing and tumorigenesis; for example, the genes identified
here may not only represent novel markers of senescence, but may also have prognostic
and/or diagnostic value in the context of tumorigenic treatment. Moreover, as there is
accumulating evidence to suggest that the induction of senescence in vivo is critical to
the efficacy of chemotherapeutic agents (Chang, Swift et al. 2002; Rebbaa, Zheng et al.
2003; Zheng, Wang et al. 2004), elucidation of the pathways critical regulating the finite
proliferative potential of normal human cells will be important for the development of
novel chemotherapeutic agents.
331
In contradiction, it has been demonstrated that senescent cells can also promote tumor
progression in a paracrine fashion. Cells undergoing replicative senescence or oncogeneinduced senescence secret growth factors, inflammatory cytokines and chemokines, and
extracellular matrix-degrading proteases that enhance the proliferation, invasion and
angiogenesis of nearby premalignant tumor cells (Campisi and d'Adda di Fagagna
2007). This was further confirmed in this present study where many metalloproteinases,
collagenases and other extra-cellular matrix degrading enzymes, secreted factors
including interleukins and growth factor were up-regulated upon senescence.
As senescent cells accumulate with age, these observations could provide an explanation
to the age-related increase in cancer incidence. In fact, senescence may be an example of
antagonistic pleiotropy, acting as a tumor suppressor mechanism in the young but
promoting tumor formation in the ederly.
8
REFERENCES
Acosta, J. C., A. O'Loghlen, et al. (2008). "Chemokine signaling via the CXCR2
receptor reinforces senescence." Cell 133(6): 1006-1018.
Adams, P. D. (2001). "Regulation of the retinoblastoma tumor suppressor protein by
cyclin/cdks." Biochim Biophys Acta 1471(3): M123-133.
Adams, P. D. (2009). "Healing and hurting: molecular mechanisms, functions, and
pathologies of cellular senescence." Mol Cell 36(1): 2-14.
Adler, A. S., M. Lin, et al. (2006). "Genetic regulators of large-scale transcriptional
signatures in cancer." Nat Genet 38(4): 421-430.
Adler, A. S., S. Sinha, et al. (2007). "Motif module map reveals enforcement of aging by
continual NF-kappaB activity." Genes Dev 21(24): 3244-3257.
Akao, Y., Y. Nakagawa, et al. (2006). "MicroRNAs 143 and 145 are possible common
onco-microRNAs in human cancers." Oncol Rep 16(4): 845-850.
Alani, R. M., A. Z. Young, et al. (2001). "Id1 regulation of cellular senescence through
transcriptional repression of p16/Ink4a." Proc Natl Acad Sci U S A 98(14):
7812-7816.
Alarcon-Vargas, D. and Z. Ronai (2002). "p53-Mdm2--the affair that never ends."
Carcinogenesis 23(4): 541-547.
Albrechtsen, N., I. Dornreiter, et al. (1999). "Maintenance of genomic integrity by p53:
complementary roles for activated and non-activated p53." Oncogene 18(53):
7706-7717.
Alcalay, M., L. Tomassoni, et al. (1998). "The promyelocytic leukemia gene product
(PML) forms stable complexes with the retinoblastoma protein." Mol Cell Biol
18(2): 1084-1093.
332
Alcorta, D. A., Y. Xiong, et al. (1996). "Involvement of the cyclin-dependent kinase
inhibitor p16 (INK4a) in replicative senescence of normal human fibroblasts."
Proc Natl Acad Sci U S A 93(24): 13742-13747.
Ali, S. H. and J. A. DeCaprio (2001). "Cellular transformation by SV40 large T antigen:
interaction with host proteins." Semin Cancer Biol 11(1): 15-23.
Alvarez-Salas, L. M., A. E. Cullinan, et al. (1998). "Inhibition of HPV-16 E6/E7
immortalization of normal keratinocytes by hairpin ribozymes." Proc Natl Acad
Sci U S A 95(3): 1189-1194.
Alwine, J. C., S. I. Reed, et al. (1977). "Characterization of the autoregulation of simian
virus 40 gene A." J Virol 24(1): 22-27.
Ambros, V. (2004). "The functions of animal microRNAs." Nature 431(7006): 350-355.
Appella, E. and C. W. Anderson (2001). "Post-translational modifications and activation
of p53 by genotoxic stresses." Eur J Biochem 268(10): 2764-2772.
Aslanian, A., P. J. Iaquinta, et al. (2004). "Repression of the Arf tumor suppressor by
E2F3 is required for normal cell cycle kinetics." Genes Dev 18(12): 1413-1422.
Augert, A., C. Payre, et al. (2009). "The M-type receptor PLA2R regulates senescence
through the p53 pathway." EMBO Rep 10(3): 271-277.
Avantaggiati, M. L., M. Carbone, et al. (1996). "The SV40 large T antigen and
adenovirus E1a oncoproteins interact with distinct isoforms of the transcriptional
co-activator, p300." EMBO J 15(9): 2236-2248.
Bammler, T., R. P. Beyer, et al. (2005). "Standardizing global gene expression analysis
between laboratories and across platforms." Nat Methods 2(5): 351-356.
Bannister, A. J., P. Zegerman, et al. (2001). "Selective recognition of methylated lysine
9 on histone H3 by the HP1 chromo domain." Nature 410(6824): 120-124.
Barbeau, D., R. Charbonneau, et al. (1994). "Functional interactions within adenovirus
E1A protein complexes." Oncogene 9(2): 359-373.
Barradas, M., E. Anderton, et al. (2009). "Histone demethylase JMJD3 contributes to
epigenetic control of INK4a/ARF by oncogenic RAS." Genes Dev 23(10): 11771182.
Bartek, J., J. Bartkova, et al. (1996). "The retinoblastoma protein pathway and the
restriction point." Curr Opin Cell Biol 8(6): 805-814.
Bartel, D. P. (2004). "MicroRNAs: genomics, biogenesis, mechanism, and function."
Cell 116(2): 281-297.
Bartkova, J., N. Rezaei, et al. (2006). "Oncogene-induced senescence is part of the
tumorigenesis barrier imposed by DNA damage checkpoints." Nature 444(7119):
633-637.
Basak, S., H. Kim, et al. (2007). "A fourth IkappaB protein within the NF-kappaB
signaling module." Cell 128(2): 369-381.
Basak, S., V. F. Shih, et al. (2008). "Generation and activation of multiple dimeric
transcription factors within the NF-kappaB signaling system." Mol Cell Biol
28(10): 3139-3150.
Bates, S., A. C. Phillips, et al. (1998). "p14ARF links the tumour suppressors RB and
p53." Nature 395(6698): 124-125.
Beausejour, C. M., A. Krtolica, et al. (2003). "Reversal of human cellular senescence:
roles of the p53 and p16 pathways." EMBO J 22(16): 4212-4222.
Bentwich, I., A. Avniel, et al. (2005). "Identification of hundreds of conserved and
nonconserved human microRNAs." Nat Genet 37(7): 766-770.
333
Bernard, D., K. Gosselin, et al. (2004). "Involvement of Rel/nuclear factor-kappaB
transcription factors in keratinocyte senescence." Cancer Res 64(2): 472-481.
Berns, K., E. M. Hijmans, et al. (2004). "A large-scale RNAi screen in human cells
identifies new components of the p53 pathway." Nature 428(6981): 431-437.
Berry, D. E., Y. Lu, et al. (1996). "Retinoblastoma protein inhibits IFN-gamma induced
apoptosis." Oncogene 12(8): 1809-1819.
Berthois, Y., N. Pourreau-Schneider, et al. (1986). "Estradiol membrane binding sites on
human breast cancer cell lines. Use of a fluorescent estradiol conjugate to
demonstrate plasma membrane binding systems." J Steroid Biochem 25(6): 963972.
Bhaumik, D., G. K. Scott, et al. (2008). "Expression of microRNA-146 suppresses NFkappaB activity with reduction of metastatic potential in breast cancer cells."
Oncogene 27(42): 5643-5647.
Bialik, S. and A. Kimchi (2004). "DAP-kinase as a target for drug design in cancer and
diseases associated with accelerated cell death." Semin Cancer Biol 14(4): 283294.
Binet, R., D. Ythier, et al. (2009). "WNT16B is a new marker of cellular senescence that
regulates p53 activity and the phosphoinositide 3-kinase/AKT pathway." Cancer
Res 69(24): 9183-9191.
Blasco, M. A. (2005). "Telomeres and human disease: ageing, cancer and beyond." Nat
Rev Genet 6(8): 611-622.
Blasco, M. A., H. W. Lee, et al. (1997). "Telomere shortening and tumor formation by
mouse cells lacking telomerase RNA." Cell 91(1): 25-34.
Boden, D., O. Pusch, et al. (2004). "Enhanced gene silencing of HIV-1 specific siRNA
using microRNA designed hairpins." Nucleic Acids Res 32(3): 1154-1158.
Bodnar, A. G., M. Ouellette, et al. (1998). "Extension of life-span by introduction of
telomerase into normal human cells." Science 279(5349): 349-352.
Bolstad, B. M., R. A. Irizarry, et al. (2003). "A comparison of normalization methods for
high density oligonucleotide array data based on variance and bias."
Bioinformatics 19(2): 185-193.
Bond, J., C. Jones, et al. (2004). "Direct evidence from siRNA-directed "knock down"
that p16(INK4a) is required for human fibroblast senescence and for limiting rasinduced epithelial cell proliferation." Exp Cell Res 292(1): 151-156.
Bonetto, F., M. Fanciulli, et al. (1999). "Interaction between the pRb2/p130 C-terminal
domain and the N-terminal portion of cyclin D3." J Cell Biochem 75(4): 698709.
Bonizzi, G., M. Bebien, et al. (2004). "Activation of IKKalpha target genes depends on
recognition of specific kappaB binding sites by RelB:p52 dimers." EMBO J
23(21): 4202-4210.
Borgdorff, V., M. E. Lleonart, et al. (2010). "Multiple microRNAs rescue from Rasinduced senescence by inhibiting p21(Waf1/Cip1)." Oncogene 29(15): 22622271.
Bos, J. L. (1989). "ras oncogenes in human cancer: a review." Cancer Res 49(17): 46824689.
Boulaire, J., A. Fotedar, et al. (2000). "The functions of the cdk-cyclin kinase inhibitor
p21WAF1." Pathol Biol (Paris) 48(3): 190-202.
334
Bours, V., G. Franzoso, et al. (1993). "The oncoprotein Bcl-3 directly transactivates
through kappa B motifs via association with DNA-binding p50B homodimers."
Cell 72(5): 729-739.
Braconi, C., N. Huang, et al. (2010). "MicroRNA-dependent regulation of DNA
methyltransferase-1 and tumor suppressor gene expression by interleukin-6 in
human malignant cholangiocytes." Hepatology 51(3): 881-890.
Braig, M., S. Lee, et al. (2005). "Oncogene-induced senescence as an initial barrier in
lymphoma development." Nature 436(7051): 660-665.
Braithwaite, A. W., B. F. Cheetham, et al. (1983). "Adenovirus-induced alterations of
the cell growth cycle: a requirement for expression of E1A but not of E1B." J
Virol 45(1): 192-199.
Brehm, A., E. A. Miska, et al. (1998). "Retinoblastoma protein recruits histone
deacetylase to repress transcription." Nature 391(6667): 597-601.
Brennecke, J., D. R. Hipfner, et al. (2003). "bantam encodes a developmentally
regulated microRNA that controls cell proliferation and regulates the
proapoptotic gene hid in Drosophila." Cell 113(1): 25-36.
Brennecke, J., A. Stark, et al. (2005). "Principles of microRNA-target recognition."
PLoS Biol 3(3): e85.
Brickman, J. M., M. Adam, et al. (1999). "Interactions between an HMG-1 protein and
members of the Rel family." Proc Natl Acad Sci U S A 96(19): 10679-10683.
Brookes, S., J. Rowe, et al. (2002). "INK4a-deficient human diploid fibroblasts are
resistant to RAS-induced senescence." EMBO J 21(12): 2936-2945.
Brown, D. R., S. Deb, et al. (1993). "The tumor suppressor p53 and the oncoprotein
simian virus 40 T antigen bind to overlapping domains on the MDM2 protein."
Mol Cell Biol 13(11): 6849-6857.
Brown, J. P., W. Wei, et al. (1997). "Bypass of senescence after disruption of
p21CIP1/WAF1 gene in normal diploid human fibroblasts." Science 277(5327):
831-834.
Bruemmer, D., F. Yin, et al. (2003). "Rapamycin inhibits E2F-dependent expression of
minichromosome maintenance proteins in vascular smooth muscle cells."
Biochem Biophys Res Commun 303(1): 251-258.
Buchkovich, K., L. A. Duffy, et al. (1989). "The retinoblastoma protein is
phosphorylated during specific phases of the cell cycle." Cell 58(6): 1097-1105.
Bulavin, D. V., C. Phillips, et al. (2004). "Inactivation of the Wip1 phosphatase inhibits
mammary tumorigenesis through p38 MAPK-mediated activation of the
p16(Ink4a)-p19(Arf) pathway." Nat Genet 36(4): 343-350.
Burkhart, B. A., D. A. Alcorta, et al. (1999). "Two posttranscriptional pathways that
regulate p21(Cip1/Waf1/Sdi1) are identified by HPV16-E6 interaction and
correlate with life span and cellular senescence." Exp Cell Res 247(1): 168-175.
Cahill, D. P., C. Lengauer, et al. (1998). "Mutations of mitotic checkpoint genes in
human cancers." Nature 392(6673): 300-303.
Calin, G. A., C. D. Dumitru, et al. (2002). "Frequent deletions and down-regulation of
micro- RNA genes miR15 and miR16 at 13q14 in chronic lymphocytic
leukemia." Proc Natl Acad Sci U S A 99(24): 15524-15529.
Calin, G. A., C. Sevignani, et al. (2004). "Human microRNA genes are frequently
located at fragile sites and genomic regions involved in cancers." Proc Natl Acad
Sci U S A 101(9): 2999-3004.
335
Campisi, J. (2003). "Cancer and ageing: rival demons?" Nat Rev Cancer 3(5): 339-349.
Campisi, J. (2005). "Senescent cells, tumor suppression, and organismal aging: good
citizens, bad neighbors." Cell 120(4): 513-522.
Campisi, J. and F. d'Adda di Fagagna (2007). "Cellular senescence: when bad things
happen to good cells." Nat Rev Mol Cell Biol 8(9): 729-740.
Canela, A., J. Martin-Caballero, et al. (2004). "Constitutive expression of tert in
thymocytes leads to increased incidence and dissemination of T-cell lymphoma
in Lck-Tert mice." Mol Cell Biol 24(10): 4275-4293.
Canela, A., E. Vera, et al. (2007). "High-throughput telomere length quantification by
FISH and its application to human population studies." Proc Natl Acad Sci U S
A 104(13): 5300-5305.
Caporossi, D. and S. Bacchetti (1990). "Definition of adenovirus type 5 functions
involved in the induction of chromosomal aberrations in human cells." J Gen
Virol 71 ( Pt 4): 801-808.
Carroll, R. B., L. Hager, et al. (1974). "Simian virus 40 T antigen binds to DNA." Proc
Natl Acad Sci U S A 71(9): 3754-3757.
Carroll, R. B., A. Samad, et al. (1988). "RNA is covalently linked to SV40 large T
antigen." Oncogene 2(5): 437-444.
Cartwright, P., H. Muller, et al. (1998). "E2F-6: a novel member of the E2F family is an
inhibitor of E2F-dependent transcription." Oncogene 17(5): 611-623.
Cawthon, R. M., K. R. Smith, et al. (2003). "Association between telomere length in
blood and mortality in people aged 60 years or older." Lancet 361(9355): 393395.
Chan, H. M., M. Krstic-Demonacos, et al. (2001). "Acetylation control of the
retinoblastoma tumour-suppressor protein." Nat Cell Biol 3(7): 667-674.
Chan, H. M., M. Narita, et al. (2005). "The p400 E1A-associated protein is a novel
component of the p53 --> p21 senescence pathway." Genes Dev 19(2): 196-201.
Chang, B. D., E. V. Broude, et al. (1999). "A senescence-like phenotype distinguishes
tumor cells that undergo terminal proliferation arrest after exposure to anticancer
agents." Cancer Res 59(15): 3761-3767.
Chang, B. D., M. E. Swift, et al. (2002). "Molecular determinants of terminal growth
arrest induced in tumor cells by a chemotherapeutic agent." Proc Natl Acad Sci
U S A 99(1): 389-394.
Chang, B. D., K. Watanabe, et al. (2000). "Effects of p21Waf1/Cip1/Sdi1 on cellular
gene expression: implications for carcinogenesis, senescence, and age-related
diseases." Proc Natl Acad Sci U S A 97(8): 4291-4296.
Chang, T. H., F. A. Ray, et al. (1997). "Disregulation of mitotic checkpoints and
regulatory proteins following acute expression of SV40 large T antigen in diploid
human cells." Oncogene 14(20): 2383-2393.
Chen, C. Z., L. Li, et al. (2004). "MicroRNAs modulate hematopoietic lineage
differentiation." Science 303(5654): 83-86.
Chen, J., P. K. Jackson, et al. (1995). "Separate domains of p21 involved in the
inhibition of Cdk kinase and PCNA." Nature 374(6520): 386-388.
Chen, P. Y. and G. Meister (2005). "microRNA-guided posttranscriptional gene
regulation." Biol Chem 386(12): 1205-1218.
336
Chen, Q. M., K. R. Prowse, et al. (2001). "Uncoupling the senescent phenotype from
telomere shortening in hydrogen peroxide-treated fibroblasts." Exp Cell Res
265(2): 294-303.
Chen, T. T. and J. Y. Wang (2000). "Establishment of irreversible growth arrest in
myogenic differentiation requires the RB LXCXE-binding function." Mol Cell
Biol 20(15): 5571-5580.
Chen, Z., L. C. Trotman, et al. (2005). "Crucial role of p53-dependent cellular
senescence in suppression of Pten-deficient tumorigenesis." Nature 436(7051):
725-730.
Chen, Z., W. Zhuo, et al. (2008). "Down-regulation of layilin, a novel hyaluronan
receptor, via RNA interference, inhibits invasion and lymphatic metastasis of
human lung A549 cells." Biotechnol Appl Biochem 50(Pt 2): 89-96.
Cheng, M., P. Olivier, et al. (1999). "The p21(Cip1) and p27(Kip1) CDK 'inhibitors' are
essential activators of cyclin D-dependent kinases in murine fibroblasts." EMBO
J 18(6): 1571-1583.
Chin, L., J. Pomerantz, et al. (1997). "Cooperative effects of INK4a and ras in
melanoma susceptibility in vivo." Genes Dev 11(21): 2822-2834.
Chondrogianni, N., F. L. Stratford, et al. (2003). "Central role of the proteasome in
senescence and survival of human fibroblasts: induction of a senescence-like
phenotype upon its inhibition and resistance to stress upon its activation." J Biol
Chem 278(30): 28026-28037.
Christoffersen, N. R., R. Shalgi, et al. (2010). "p53-independent upregulation of miR34a during oncogene-induced senescence represses MYC." Cell Death Differ
17(2): 236-245.
Chuang, Y. T., L. W. Fang, et al. (2008). "The tumor suppressor death-associated
protein kinase targets to TCR-stimulated NF-kappa B activation." J Immunol
180(5): 3238-3249.
Chung, E. J., Y. K. Sung, et al. (2002). "Gene expression profile analysis in human
hepatocellular carcinoma by cDNA microarray." Mol Cells 14(3): 382-387.
Classon, M. and N. Dyson (2001). "p107 and p130: versatile proteins with interesting
pockets." Exp Cell Res 264(1): 135-147.
Classon, M. and E. Harlow (2002). "The retinoblastoma tumour suppressor in
development and cancer." Nat Rev Cancer 2(12): 910-917.
Claudio, P. P., A. De Luca, et al. (1996). "Functional analysis of pRb2/p130 interaction
with cyclins." Cancer Res 56(9): 2003-2008.
Cleveland, J. L. and C. J. Sherr (2004). "Antagonism of Myc functions by Arf." Cancer
Cell 6(4): 309-311.
Cobrinik, D., S. F. Dowdy, et al. (1992). "The retinoblastoma protein and the regulation
of cell cycling." Trends Biochem Sci 17(8): 312-315.
Collado, M., M. A. Blasco, et al. (2007). "Cellular senescence in cancer and aging." Cell
130(2): 223-233.
Collado, M., J. Gil, et al. (2005). "Tumour biology: senescence in premalignant
tumours." Nature 436(7051): 642.
Collado, M. and M. Serrano (2010). "Senescence in tumours: evidence from mice and
humans." Nat Rev Cancer 10(1): 51-57.
Collins, K. and J. R. Mitchell (2002). "Telomerase in the human organism." Oncogene
21(4): 564-579.
337
Colman, M. S., C. A. Afshari, et al. (2000). "Regulation of p53 stability and activity in
response to genotoxic stress." Mutat Res 462(2-3): 179-188.
Condorelli, G. and A. Giordano (1997). "Synergistic role of E1A-binding proteins and
tissue-specific transcription factors in differentiation." J Cell Biochem 67(4):
423-431.
Condorelli, G. L., U. Testa, et al. (1995). "Modulation of retinoblastoma gene in normal
adult hematopoiesis: peak expression and functional role in advanced erythroid
differentiation." Proc Natl Acad Sci U S A 92(11): 4808-4812.
Conzen, S. D. and C. N. Cole (1995). "The three transforming regions of SV40 T
antigen are required for immortalization of primary mouse embryo fibroblasts."
Oncogene 11(11): 2295-2302.
Coppe, J. P., C. K. Patil, et al. (2008). "Senescence-associated secretory phenotypes
reveal cell-nonautonomous functions of oncogenic RAS and the p53 tumor
suppressor." PLoS Biol 6(12): 2853-2868.
Core, N., F. Joly, et al. (2004). "Disruption of E2F signaling suppresses the INK4ainduced proliferative defect in M33-deficient mice." Oncogene 23(46): 76607668.
Corney, D. C., C. I. Hwang, et al. (2010). "Frequent downregulation of miR-34 family in
human ovarian cancers." Clin Cancer Res 16(4): 1119-1128.
Cosme-Blanco, W., M. F. Shen, et al. (2007). "Telomere dysfunction suppresses
spontaneous tumorigenesis in vivo by initiating p53-dependent cellular
senescence." EMBO Rep 8(5): 497-503.
Cotsiki, M., R. L. Lock, et al. (2004). "Simian virus 40 large T antigen targets the
spindle assembly checkpoint protein Bub1." Proc Natl Acad Sci U S A 101(4):
947-952.
Counter, C. M., M. Meyerson, et al. (1998). "Telomerase activity is restored in human
cells by ectopic expression of hTERT (hEST2), the catalytic subunit of
telomerase." Oncogene 16(9): 1217-1222.
Courtois-Cox, S., S. M. Genther Williams, et al. (2006). "A negative feedback signaling
network underlies oncogene-induced senescence." Cancer Cell 10(6): 459-472.
Courtois-Cox, S., S. L. Jones, et al. (2008). "Many roads lead to oncogene-induced
senescence." Oncogene 27(20): 2801-2809.
Cristofalo, V. J. and M. Tresini (1998). "Defects in signal transduction during replicative
senescence of diploid human fibroblasts in vitro." Aging (Milano) 10(2): 151152.
Cui, J. W., Y. J. Li, et al. (2007). "Retroviral insertional activation of the Fli-3 locus in
erythroleukemias encoding a cluster of microRNAs that convert Epo-induced
differentiation to proliferation." Blood 110(7): 2631-2640.
Cziepluch, C., S. Lampel, et al. (2000). "H-1 parvovirus-associated replication bodies: a
distinct virus-induced nuclear structure." J Virol 74(10): 4807-4815.
d'Adda di Fagagna, F., P. M. Reaper, et al. (2003). "A DNA damage checkpoint
response in telomere-initiated senescence." Nature 426(6963): 194-198.
d'Adda di Fagagna, F., S. H. Teo, et al. (2004). "Functional links between telomeres and
proteins of the DNA-damage response." Genes Dev 18(15): 1781-1799.
Dahiya, A., M. R. Gavin, et al. (2000). "Role of the LXCXE binding site in Rb
function." Mol Cell Biol 20(18): 6799-6805.
338
Dankort, D., E. Filenova, et al. (2007). "A new mouse model to explore the initiation,
progression, and therapy of BRAFV600E-induced lung tumors." Genes Dev
21(4): 379-384.
Davies, H., C. Hunter, et al. (2005). "Somatic mutations of the protein kinase gene
family in human lung cancer." Cancer Res 65(17): 7591-7595.
Davies, R., R. Hicks, et al. (1993). "Human papillomavirus type 16 E7 associates with a
histone H1 kinase and with p107 through sequences necessary for
transformation." J Virol 67(5): 2521-2528.
de Bruin, A., B. Maiti, et al. (2003). "Identification and characterization of E2F7, a
novel mammalian E2F family member capable of blocking cellular
proliferation." J Biol Chem 278(43): 42041-42049.
de Lange, T. (2005). "Shelterin: the protein complex that shapes and safeguards human
telomeres." Genes Dev 19(18): 2100-2110.
de Stanchina, E., M. E. McCurrach, et al. (1998). "E1A signaling to p53 involves the
p19(ARF) tumor suppressor." Genes Dev 12(15): 2434-2442.
Debacq, C., J. M. Heraud, et al. (2005). "Reduced cell turnover in lymphocytic monkeys
infected by human T-lymphotropic virus type 1." Oncogene 24(51): 7514-7523.
DeCaprio, J. A., J. W. Ludlow, et al. (1988). "SV40 large tumor antigen forms a specific
complex with the product of the retinoblastoma susceptibility gene." Cell 54(2):
275-283.
DeCaprio, J. A., J. W. Ludlow, et al. (1989). "The product of the retinoblastoma
susceptibility gene has properties of a cell cycle regulatory element." Cell 58(6):
1085-1095.
DeGregori, J., G. Leone, et al. (1997). "Distinct roles for E2F proteins in cell growth
control and apoptosis." Proc Natl Acad Sci U S A 94(14): 7245-7250.
Deiss, L. P., E. Feinstein, et al. (1995). "Identification of a novel serine/threonine kinase
and a novel 15-kD protein as potential mediators of the gamma interferoninduced cell death." Genes Dev 9(1): 15-30.
DeLeo, A. B., G. Jay, et al. (1979). "Detection of a transformation-related antigen in
chemically induced sarcomas and other transformed cells of the mouse." Proc
Natl Acad Sci U S A 76(5): 2420-2424.
Deppert, W., M. Haug, et al. (1987). "Modulation of p53 protein expression during
cellular transformation with simian virus 40." Mol Cell Biol 7(12): 4453-4463.
Der, C. J., T. G. Krontiris, et al. (1982). "Transforming genes of human bladder and lung
carcinoma cell lines are homologous to the ras genes of Harvey and Kirsten
sarcoma viruses." Proc Natl Acad Sci U S A 79(11): 3637-3640.
Derynck, R., R. J. Akhurst, et al. (2001). "TGF-beta signaling in tumor suppression and
cancer progression." Nat Genet 29(2): 117-129.
Dews, M., A. Homayouni, et al. (2006). "Augmentation of tumor angiogenesis by a
Myc-activated microRNA cluster." Nat Genet 38(9): 1060-1065.
Di Leonardo, A., S. P. Linke, et al. (1994). "DNA damage triggers a prolonged p53dependent G1 arrest and long-term induction of Cip1 in normal human
fibroblasts." Genes Dev 8(21): 2540-2551.
Di Micco, P., M. Amitrano, et al. (2006). "Molecular and clinical conditions associated
with venous thromboembolism in oncological patients." Exp Oncol 28(3): 194197.
339
Dimri, G. P., E. Hara, et al. (1994). "Regulation of two E2F-related genes in
presenescent and senescent human fibroblasts." J Biol Chem 269(23): 1618016186.
Dimri, G. P., K. Itahana, et al. (2000). "Regulation of a senescence checkpoint response
by the E2F1 transcription factor and p14(ARF) tumor suppressor." Mol Cell Biol
20(1): 273-285.
Dimri, G. P., X. Lee, et al. (1995). "A biomarker that identifies senescent human cells in
culture and in aging skin in vivo." Proc Natl Acad Sci U S A 92(20): 9363-9367.
Dirac, A. M. and R. Bernards (2003). "Reversal of senescence in mouse fibroblasts
through lentiviral suppression of p53." J Biol Chem 278(14): 11731-11734.
Dobrzanski, P., R. P. Ryseck, et al. (1995). "Specific inhibition of RelB/p52
transcriptional activity by the C-terminal domain of p100." Oncogene 10(5):
1003-1007.
Dodeller, F., M. Gottar, et al. (2008). "The lysosomal transmembrane protein 9B
regulates the activity of inflammatory signaling pathways." J Biol Chem
283(31): 21487-21494.
Donehower, L. A., L. A. Godley, et al. (1995). "Deficiency of p53 accelerates mammary
tumorigenesis in Wnt-1 transgenic mice and promotes chromosomal instability."
Genes Dev 9(7): 882-895.
Donehower, L. A., M. Harvey, et al. (1992). "Mice deficient for p53 are
developmentally normal but susceptible to spontaneous tumours." Nature
356(6366): 215-221.
Dornan, D., I. Wertz, et al. (2004). "The ubiquitin ligase COP1 is a critical negative
regulator of p53." Nature 429(6987): 86-92.
Dorsman, J. C., B. M. Hagmeyer, et al. (1995). "The N-terminal region of the
adenovirus type 5 E1A proteins can repress expression of cellular genes via two
distinct but overlapping domains." J Virol 69(5): 2962-2967.
Doyle, S. L. and L. A. O'Neill (2006). "Toll-like receptors: from the discovery of
NFkappaB to new insights into transcriptional regulations in innate immunity."
Biochem Pharmacol 72(9): 1102-1113.
Draviam, V. M., F. Stegmeier, et al. (2007). "A functional genomic screen identifies a
role for TAO1 kinase in spindle-checkpoint signalling." Nat Cell Biol 9(5): 556564.
Duensing, S. and K. Munger (2002). "The human papillomavirus type 16 E6 and E7
oncoproteins independently induce numerical and structural chromosome
instability." Cancer Res 62(23): 7075-7082.
Dumble, M., C. Gatza, et al. (2004). "Insights into aging obtained from p53 mutant
mouse models." Ann N Y Acad Sci 1019: 171-177.
Dumble, M., L. Moore, et al. (2007). "The impact of altered p53 dosage on
hematopoietic stem cell dynamics during aging." Blood 109(4): 1736-1742.
Dunaief, J. L., B. E. Strober, et al. (1994). "The retinoblastoma protein and BRG1 form
a complex and cooperate to induce cell cycle arrest." Cell 79(1): 119-130.
Dyson, N., K. Buchkovich, et al. (1989). "The cellular 107K protein that binds to
adenovirus E1A also associates with the large T antigens of SV40 and JC virus."
Cell 58(2): 249-255.
Dyson, N. and E. Harlow (1992). "Adenovirus E1A targets key regulators of cell
proliferation." Cancer Surv 12: 161-195.
340
Eckner, R. (1996). "p300 and CBP as transcriptional regulators and targets of oncogenic
events." Biol Chem 377(11): 685-688.
Eckner, R., J. W. Ludlow, et al. (1996). "Association of p300 and CBP with simian virus
40 large T antigen." Mol Cell Biol 16(7): 3454-3464.
Eklov, S., K. Funa, et al. (1993). "Lack of the latent transforming growth factor beta
binding protein in malignant, but not benign prostatic tissue." Cancer Res 53(13):
3193-3197.
el-Deiry, W. S. (1998). "Regulation of p53 downstream genes." Semin Cancer Biol 8(5):
345-357.
el-Deiry, W. S., T. Tokino, et al. (1993). "WAF1, a potential mediator of p53 tumor
suppression." Cell 75(4): 817-825.
Elenbaas, B., L. Spirio, et al. (2001). "Human breast cancer cells generated by oncogenic
transformation of primary mammary epithelial cells." Genes Dev 15(1): 50-65.
Esau, C., S. Davis, et al. (2006). "miR-122 regulation of lipid metabolism revealed by in
vivo antisense targeting." Cell Metab 3(2): 87-98.
Esau, C. C. (2008). "Inhibition of microRNA with antisense oligonucleotides." Methods
44(1): 55-60.
Esteller, M., S. Gonzalez, et al. (2001). "K-ras and p16 aberrations confer poor
prognosis in human colorectal cancer." J Clin Oncol 19(2): 299-304.
Esteller, M., S. Tortola, et al. (2000). "Hypermethylation-associated inactivation of
p14(ARF) is independent of p16(INK4a) methylation and p53 mutational status."
Cancer Res 60(1): 129-133.
Evans, T., E. T. Rosenthal, et al. (1983). "Cyclin: a protein specified by maternal mRNA
in sea urchin eggs that is destroyed at each cleavage division." Cell 33(2): 389396.
Ewen, M. E., J. W. Ludlow, et al. (1989). "An N-terminal transformation-governing
sequence of SV40 large T antigen contributes to the binding of both p110Rb and
a second cellular protein, p120." Cell 58(2): 257-267.
Fabani, M. M. and M. J. Gait (2008). "miR-122 targeting with LNA/2'-O-methyl
oligonucleotide mixmers, peptide nucleic acids (PNA), and PNA-peptide
conjugates." RNA 14(2): 336-346.
Faha, B., M. E. Ewen, et al. (1992). "Interaction between human cyclin A and
adenovirus E1A-associated p107 protein." Science 255(5040): 87-90.
Fakharzadeh, S. S., S. P. Trusko, et al. (1991). "Tumorigenic potential associated with
enhanced expression of a gene that is amplified in a mouse tumor cell line."
EMBO J 10(6): 1565-1569.
Ferbeyre, G., E. de Stanchina, et al. (2002). "Oncogenic ras and p53 cooperate to induce
cellular senescence." Mol Cell Biol 22(10): 3497-3508.
Filippova, M., H. Song, et al. (2002). "The human papillomavirus 16 E6 protein binds to
tumor necrosis factor (TNF) R1 and protects cells from TNF-induced apoptosis."
J Biol Chem 277(24): 21730-21739.
Finkel, T., M. Serrano, et al. (2007). "The common biology of cancer and ageing."
Nature 448(7155): 767-774.
Flemington, E. K., S. H. Speck, et al. (1993). "E2F-1-mediated transactivation is
inhibited by complex formation with the retinoblastoma susceptibility gene
product." Proc Natl Acad Sci U S A 90(15): 6914-6918.
341
Frame, S. and A. Balmain (2000). "Integration of positive and negative growth signals
during ras pathway activation in vivo." Curr Opin Genet Dev 10(1): 106-113.
Franzoso, G., V. Bours, et al. (1992). "The candidate oncoprotein Bcl-3 is an antagonist
of p50/NF-kappa B-mediated inhibition." Nature 359(6393): 339-342.
Friedman, S. R., W. de Jong, et al. (2007). "Harm reduction theory: users' culture,
micro-social indigenous harm reduction, and the self-organization and outsideorganizing of users' groups." Int J Drug Policy 18(2): 107-117.
Frippiat, C., Q. M. Chen, et al. (2001). "Subcytotoxic H2O2 stress triggers a release of
transforming growth factor-beta 1, which induces biomarkers of cellular
senescence of human diploid fibroblasts." J Biol Chem 276(4): 2531-2537.
Frippiat, C., J. Dewelle, et al. (2002). "Signal transduction in H2O2-induced senescencelike phenotype in human diploid fibroblasts." Free Radic Biol Med 33(10): 13341346.
Frolov, M. V. and N. J. Dyson (2004). "Molecular mechanisms of E2F-dependent
activation and pRB-mediated repression." J Cell Sci 117(Pt 11): 2173-2181.
Fujita, T., G. P. Nolan, et al. (1993). "The candidate proto-oncogene bcl-3 encodes a
transcriptional coactivator that activates through NF-kappa B p50 homodimers."
Genes Dev 7(7B): 1354-1363.
Fukasawa, K. and G. F. Vande Woude (1997). "Synergy between the Mos/mitogenactivated protein kinase pathway and loss of p53 function in transformation and
chromosome instability." Mol Cell Biol 17(1): 506-518.
Gage, J. R., C. Meyers, et al. (1990). "The E7 proteins of the nononcogenic human
papillomavirus type 6b (HPV-6b) and of the oncogenic HPV-16 differ in
retinoblastoma protein binding and other properties." J Virol 64(2): 723-730.
Gao, C. P., Z. Y. Zhang, et al. (2010). "[Reduced expression of miR-218 and its
significance in gastric cancer]." Zhonghua Zhong Liu Za Zhi 32(4): 249-252.
Gao, F., J. F. Ponte, et al. (2009). "hBub1 negatively regulates p53 mediated early cell
death upon mitotic checkpoint activation." Cancer Biol Ther 8(7): 548-556.
Garcia-Cao, I., M. Garcia-Cao, et al. (2006). "Increased p53 activity does not accelerate
telomere-driven ageing." EMBO Rep 7(5): 546-552.
Gartel, A. L., M. S. Serfas, et al. (1996). "p21 (WAF1/CIP1) expression is induced in
newly nondividing cells in diverse epithelia and during differentiation of the
Caco-2 intestinal cell line." Exp Cell Res 227(2): 171-181.
Gaubatz, S., J. G. Wood, et al. (1998). "Unusual proliferation arrest and transcriptional
control properties of a newly discovered E2F family member, E2F-6." Proc Natl
Acad Sci U S A 95(16): 9190-9195.
Gewin, L. and D. A. Galloway (2001). "E box-dependent activation of telomerase by
human papillomavirus type 16 E6 does not require induction of c-myc." J Virol
75(15): 7198-7201.
Gil, J., D. Bernard, et al. (2004). "Polycomb CBX7 has a unifying role in cellular
lifespan." Nat Cell Biol 6(1): 67-72.
Gilinger, G. and J. C. Alwine (1993). "Transcriptional activation by simian virus 40
large T antigen: requirements for simple promoter structures containing either
TATA or initiator elements with variable upstream factor binding sites." J Virol
67(11): 6682-6688.
342
Gillison, M. L., W. M. Koch, et al. (2000). "Evidence for a causal association between
human papillomavirus and a subset of head and neck cancers." J Natl Cancer Inst
92(9): 709-720.
Gilmore, T. D. (2006). "Introduction to NF-kappaB: players, pathways, perspectives."
Oncogene 25(51): 6680-6684.
Gjoerup, O. V., J. Wu, et al. (2007). "Surveillance mechanism linking Bub1 loss to the
p53 pathway." Proc Natl Acad Sci U S A 104(20): 8334-8339.
Gonos, E. S., J. S. Burns, et al. (1996). "Rat embryo fibroblasts immortalized with
simian virus 40 large T antigen undergo senescence upon its inactivation." Mol
Cell Biol 16(9): 5127-5138.
Gonzalez-Suarez, E., C. Geserick, et al. (2005). "Antagonistic effects of telomerase on
cancer and aging in K5-mTert transgenic mice." Oncogene 24(13): 2256-2270.
Gonzalez-Suarez, E., E. Samper, et al. (2000). "Telomerase-deficient mice with short
telomeres are resistant to skin tumorigenesis." Nat Genet 26(1): 114-117.
Goodman, R. H. and S. Smolik (2000). "CBP/p300 in cell growth, transformation, and
development." Genes Dev 14(13): 1553-1577.
Gorman, S. D. and V. J. Cristofalo (1985). "Reinitiation of cellular DNA synthesis in
BrdU-selected nondividing senescent WI-38 cells by simian virus 40 infection."
J Cell Physiol 125(1): 122-126.
Gostissa, M., A. Hengstermann, et al. (1999). "Activation of p53 by conjugation to the
ubiquitin-like protein SUMO-1." EMBO J 18(22): 6462-6471.
Greider, C. W. and E. H. Blackburn (1985). "Identification of a specific telomere
terminal transferase activity in Tetrahymena extracts." Cell 43(2 Pt 1): 405-413.
Grobet, L., L. J. Martin, et al. (1997). "A deletion in the bovine myostatin gene causes
the double-muscled phenotype in cattle." Nat Genet 17(1): 71-74.
Grosch, S., I. Tegeder, et al. (2001). "COX-2 independent induction of cell cycle arrest
and apoptosis in colon cancer cells by the selective COX-2 inhibitor celecoxib."
FASEB J 15(14): 2742-2744.
Grossman, S. R., M. Perez, et al. (1998). "p300/MDM2 complexes participate in
MDM2-mediated p53 degradation." Mol Cell 2(4): 405-415.
Groth, A., J. D. Weber, et al. (2000). "Oncogenic Ras induces p19ARF and growth
arrest in mouse embryo fibroblasts lacking p21Cip1 and p27Kip1 without
activating cyclin D-dependent kinases." J Biol Chem 275(35): 27473-27480.
Gruber, A. D. and B. U. Pauli (1999). "Tumorigenicity of human breast cancer is
associated with loss of the Ca2+-activated chloride channel CLCA2." Cancer
Res 59(21): 5488-5491.
Gruda, M. C., J. M. Zabolotny, et al. (1993). "Transcriptional activation by simian virus
40 large T antigen: interactions with multiple components of the transcription
complex." Mol Cell Biol 13(2): 961-969.
Gruis, N. A., J. Weaver-Feldhaus, et al. (1995). "Genetic evidence in melanoma and
bladder cancers that p16 and p53 function in separate pathways of tumor
suppression." Am J Pathol 146(5): 1199-1206.
Gu, W., J. W. Schneider, et al. (1993). "Interaction of myogenic factors and the
retinoblastoma protein mediates muscle cell commitment and differentiation."
Cell 72(3): 309-324.
Guimaraes, D. P. and P. Hainaut (2002). "TP53: a key gene in human cancer."
Biochimie 84(1): 83-93.
343
Guled, M., L. Lahti, et al. (2009). "CDKN2A, NF2, and JUN are dysregulated among
other genes by miRNAs in malignant mesothelioma -A miRNA microarray
analysis." Genes Chromosomes Cancer 48(7): 615-623.
Haas-Kogan, D. A., S. C. Kogan, et al. (1995). "Inhibition of apoptosis by the
retinoblastoma gene product." EMBO J 14(3): 461-472.
Haddad, M. M., W. Xu, et al. (1999). "Activation of a cAMP pathway and induction of
melanogenesis correlate with association of p16(INK4) and p27(KIP1) to CDKs,
loss of E2F-binding activity, and premature senescence of human melanocytes."
Exp Cell Res 253(2): 561-572.
Hahn, W. C. and R. A. Weinberg (2002). "Rules for making human tumor cells." N Engl
J Med 347(20): 1593-1603.
Hainaut, P., T. Soussi, et al. (1997). "Database of p53 gene somatic mutations in human
tumors and cell lines: updated compilation and future prospects." Nucleic Acids
Res 25(1): 151-157.
Halazonetis, T. D., V. G. Gorgoulis, et al. (2008). "An oncogene-induced DNA damage
model for cancer development." Science 319(5868): 1352-1355.
Hames, R. S. and A. M. Fry (2002). "Alternative splice variants of the human
centrosome kinase Nek2 exhibit distinct patterns of expression in mitosis."
Biochem J 361(Pt 1): 77-85.
Hanahan, D. and R. A. Weinberg (2000). "The hallmarks of cancer." Cell 100(1): 57-70.
Hanissian, S. H., B. Teng, et al. (2005). "Regulation of myeloid leukemia factor-1
interacting protein (MLF1IP) expression in glioblastoma." Brain Res 1047(1):
56-64.
Hannon, G. J., D. Demetrick, et al. (1993). "Isolation of the Rb-related p130 through its
interaction with CDK2 and cyclins." Genes Dev 7(12A): 2378-2391.
Hansen, U., D. G. Tenen, et al. (1981). "T antigen repression of SV40 early transcription
from two promoters." Cell 27(3 Pt 2): 603-613.
Hara, E., R. Smith, et al. (1996). "Regulation of p16CDKN2 expression and its
implications for cell immortalization and senescence." Mol Cell Biol 16(3): 859867.
Hardy, K., L. Mansfield, et al. (2005). "Transcriptional networks and cellular senescence
in human mammary fibroblasts." Mol Biol Cell 16(2): 943-953.
Harley, C. B., A. B. Futcher, et al. (1990). "Telomeres shorten during ageing of human
fibroblasts." Nature 345(6274): 458-460.
Harlow, E., P. Whyte, et al. (1986). "Association of adenovirus early-region 1A proteins
with cellular polypeptides." Mol Cell Biol 6(5): 1579-1589.
Harper, J. W., G. R. Adami, et al. (1993). "The p21 Cdk-interacting protein Cip1 is a
potent inhibitor of G1 cyclin-dependent kinases." Cell 75(4): 805-816.
Harvat, B. L., A. Wang, et al. (1998). "Up-regulation of p27Kip1, p21WAF1/Cip1 and
p16Ink4a is associated with, but not sufficient for, induction of squamous
differentiation." J Cell Sci 111 ( Pt 9): 1185-1196.
Harvey, M., M. J. McArthur, et al. (1993). "Spontaneous and carcinogen-induced
tumorigenesis in p53-deficient mice." Nat Genet 5(3): 225-229.
Hayden, M. S., A. P. West, et al. (2006). "NF-kappaB and the immune response."
Oncogene 25(51): 6758-6780.
Hayflick, L. and P. S. Moorhead (1961). "The serial cultivation of human diploid cell
strains." Exp Cell Res 25: 585-621.
344
Hayward, D. G., R. B. Clarke, et al. (2004). "The centrosomal kinase Nek2 displays
elevated levels of protein expression in human breast cancer." Cancer Res
64(20): 7370-7376.
Hayward, D. G. and A. M. Fry (2006). "Nek2 kinase in chromosome instability and
cancer." Cancer Lett 237(2): 155-166.
He, L., X. He, et al. (2007). "A microRNA component of the p53 tumour suppressor
network." Nature 447(7148): 1130-1134.
He, L., J. M. Thomson, et al. (2005). "A microRNA polycistron as a potential human
oncogene." Nature 435(7043): 828-833.
He, S., B. L. Cook, et al. (2000). "E2F is required to prevent inappropriate S-phase entry
of mammalian cells." Mol Cell Biol 20(1): 363-371.
Heinrichs, S. and W. Deppert (2003). "Apoptosis or growth arrest: modulation of the
cellular response to p53 by proliferative signals." Oncogene 22(4): 555-571.
Helenius, M., M. Hanninen, et al. (1996). "Aging-induced up-regulation of nuclear
binding activities of oxidative stress responsive NF-kB transcription factor in
mouse cardiac muscle." J Mol Cell Cardiol 28(3): 487-498.
Helin, K., E. Harlow, et al. (1993). "Inhibition of E2F-1 transactivation by direct binding
of the retinoblastoma protein." Mol Cell Biol 13(10): 6501-6508.
Helt, A. M., J. O. Funk, et al. (2002). "Inactivation of both the retinoblastoma tumor
suppressor and p21 by the human papillomavirus type 16 E7 oncoprotein is
necessary to inhibit cell cycle arrest in human epithelial cells." J Virol 76(20):
10559-10568.
Hemann, M. T., K. L. Rudolph, et al. (2001). "Telomere dysfunction triggers
developmentally regulated germ cell apoptosis." Mol Biol Cell 12(7): 20232030.
Heneghan, H. M., N. Miller, et al. (2010). "Systemic miRNA-195 differentiates breast
cancer from other malignancies and is a potential biomarker for detecting
noninvasive and early stage disease." Oncologist 15(7): 673-682.
Henning, W., G. Rohaly, et al. (1997). "MDM2 is a target of simian virus 40 in cellular
transformation and during lytic infection." J Virol 71(10): 7609-7618.
Herbig, U., W. A. Jobling, et al. (2004). "Telomere shortening triggers senescence of
human cells through a pathway involving ATM, p53, and p21(CIP1), but not
p16(INK4a)." Mol Cell 14(4): 501-513.
Hilleman, M. R. (1998). "Discovery of simian virus 40 (SV40) and its relationship to
poliomyelitis virus vaccines." Dev Biol Stand 94: 183-190.
Hoggard, N., B. Brintnell, et al. (1995). "Allelic imbalance on chromosome 1 in human
breast cancer. II. Microsatellite repeat analysis." Genes Chromosomes Cancer
12(1): 24-31.
Holland, E. A., S. C. Beaton, et al. (1995). "Analysis of the p16 gene, CDKN2, in 17
Australian melanoma kindreds." Oncogene 11(11): 2289-2294.
Honda, R. and H. Yasuda (1999). "Association of p19(ARF) with Mdm2 inhibits
ubiquitin ligase activity of Mdm2 for tumor suppressor p53." EMBO J 18(1): 2227.
Hsieh, J. K., D. Yap, et al. (2002). "Novel function of the cyclin A binding site of E2F in
regulating p53-induced apoptosis in response to DNA damage." Mol Cell Biol
22(1): 78-93.
345
Hu, G., J. Kim, et al. (2009). "A genome-wide RNAi screen identifies a new
transcriptional module required for self-renewal." Genes Dev 23(7): 837-848.
Hu, Q. J., N. Dyson, et al. (1990). "The regions of the retinoblastoma protein needed for
binding to adenovirus E1A or SV40 large T antigen are common sites for
mutations." EMBO J 9(4): 1147-1155.
Huang, S., N. P. Wang, et al. (1990). "Two distinct and frequently mutated regions of
retinoblastoma protein are required for binding to SV40 T antigen." EMBO J
9(6): 1815-1822.
Hui, A. B., M. Lenarduzzi, et al. (2010). "Comprehensive MicroRNA profiling for head
and neck squamous cell carcinomas." Clin Cancer Res 16(4): 1129-1139.
Huibregtse, J. M., M. Scheffner, et al. (1991). "A cellular protein mediates association of
p53 with the E6 oncoprotein of human papillomavirus types 16 or 18." EMBO J
10(13): 4129-4135.
Hurst, D. R., Y. Xie, et al. (2009). "Multiple forms of BRMS1 are differentially
expressed in the MCF10 isogenic breast cancer progression model." Clin Exp
Metastasis 26(2): 89-96.
Hussussian, C. J., J. P. Struewing, et al. (1994). "Germline p16 mutations in familial
melanoma." Nat Genet 8(1): 15-21.
Iavarone, A., P. Garg, et al. (1994). "The helix-loop-helix protein Id-2 enhances cell
proliferation and binds to the retinoblastoma protein." Genes Dev 8(11): 12701284.
Ikram, Z., T. Norton, et al. (1994). "The biological clock that measures the mitotic lifespan of mouse embryo fibroblasts continues to function in the presence of simian
virus 40 large tumor antigen." Proc Natl Acad Sci U S A 91(14): 6448-6452.
Imamura, T., H. Izumi, et al. (2001). "Interaction with p53 enhances binding of
cisplatin-modified DNA by high mobility group 1 protein." J Biol Chem
276(10): 7534-7540.
Inoue, J., L. D. Kerr, et al. (1992). "I kappa B gamma, a 70 kd protein identical to the Cterminal half of p110 NF-kappa B: a new member of the I kappa B family." Cell
68(6): 1109-1120.
Inoue, K., R. Wen, et al. (2000). "Disruption of the ARF transcriptional activator DMP1
facilitates cell immortalization, Ras transformation, and tumorigenesis." Genes
Dev 14(14): 1797-1809.
Inoue, K., F. Zindy, et al. (2001). "Dmp1 is haplo-insufficient for tumor suppression and
modifies the frequencies of Arf and p53 mutations in Myc-induced lymphomas."
Genes Dev 15(22): 2934-2939.
Iorio, M. V. and C. M. Croce (2009). "MicroRNAs in cancer: small molecules with a
huge impact." J Clin Oncol 27(34): 5848-5856.
Iorio, M. V., M. Ferracin, et al. (2005). "MicroRNA gene expression deregulation in
human breast cancer." Cancer Res 65(16): 7065-7070.
Irizarry, R. A., B. Hobbs, et al. (2003). "Exploration, normalization, and summaries of
high density oligonucleotide array probe level data." Biostatistics 4(2): 249-264.
Irizarry, R. A., D. Warren, et al. (2005). "Multiple-laboratory comparison of microarray
platforms." Nat Methods 2(5): 345-350.
Irwin, M. S. and W. G. Kaelin (2001). "p53 family update: p73 and p63 develop their
own identities." Cell Growth Differ 12(7): 337-349.
346
Ishikawa, F. (1997). "Regulation mechanisms of mammalian telomerase. A review."
Biochemistry (Mosc) 62(11): 1332-1337.
Itahana, K., Y. Zou, et al. (2003). "Control of the replicative life span of human
fibroblasts by p16 and the polycomb protein Bmi-1." Mol Cell Biol 23(1): 389401.
Ito, K., S. Adachi, et al. (2001). "N-Terminally extended human ubiquitin-conjugating
enzymes (E2s) mediate the ubiquitination of RING-finger proteins, ARA54 and
RNF8." Eur J Biochem 268(9): 2725-2732.
Jackson, A. L., S. R. Bartz, et al. (2003). "Expression profiling reveals off-target gene
regulation by RNAi." Nat Biotechnol 21(6): 635-637.
Jacobs, J. J. and T. de Lange (2004). "Significant role for p16INK4a in p53-independent
telomere-directed senescence." Curr Biol 14(24): 2302-2308.
Jacobs, J. J., P. Keblusek, et al. (2000). "Senescence bypass screen identifies TBX2,
which represses Cdkn2a (p19(ARF)) and is amplified in a subset of human breast
cancers." Nat Genet 26(3): 291-299.
Jacobs, J. J., K. Kieboom, et al. (1999). "The oncogene and Polycomb-group gene bmi-1
regulates cell proliferation and senescence through the ink4a locus." Nature
397(6715): 164-168.
Jacobs, M. D. and S. C. Harrison (1998). "Structure of an IkappaBalpha/NF-kappaB
complex." Cell 95(6): 749-758.
Jarrard, D. F., S. Sarkar, et al. (1999). "p16/pRb pathway alterations are required for
bypassing senescence in human prostate epithelial cells." Cancer Res 59(12):
2957-2964.
Jat, P. S. and P. A. Sharp (1989). "Cell lines established by a temperature-sensitive
simian virus 40 large-T-antigen gene are growth restricted at the nonpermissive
temperature." Mol Cell Biol 9(4): 1672-1681.
Jayaraman, L., N. C. Moorthy, et al. (1998). "High mobility group protein-1 (HMG-1) is
a unique activator of p53." Genes Dev 12(4): 462-472.
Johnson, L., D. Greenbaum, et al. (1997). "K-ras is an essential gene in the mouse with
partial functional overlap with N-ras." Genes Dev 11(19): 2468-2481.
Johnson, S. M., H. Grosshans, et al. (2005). "RAS is regulated by the let-7 microRNA
family." Cell 120(5): 635-647.
Jones, S. N., A. E. Roe, et al. (1995). "Rescue of embryonic lethality in Mdm2-deficient
mice by absence of p53." Nature 378(6553): 206-208.
Kaghad, M., H. Bonnet, et al. (1997). "Monoallelically expressed gene related to p53 at
1p36, a region frequently deleted in neuroblastoma and other human cancers."
Cell 90(4): 809-819.
Kalderon, D., W. D. Richardson, et al. (1984). "Sequence requirements for nuclear
location of simian virus 40 large-T antigen." Nature 311(5981): 33-38.
Kamijo, T., S. Bodner, et al. (1999). "Tumor spectrum in ARF-deficient mice." Cancer
Res 59(9): 2217-2222.
Kamijo, T., J. D. Weber, et al. (1998). "Functional and physical interactions of the ARF
tumor suppressor with p53 and Mdm2." Proc Natl Acad Sci U S A 95(14): 82928297.
Kamijo, T., F. Zindy, et al. (1997). "Tumor suppression at the mouse INK4a locus
mediated by the alternative reading frame product p19ARF." Cell 91(5): 649659.
347
Kan, T., F. Sato, et al. (2009). "The miR-106b-25 polycistron, activated by genomic
amplification, functions as an oncogene by suppressing p21 and Bim."
Gastroenterology 136(5): 1689-1700.
Kanehira, M., Y. Harada, et al. (2007). "Involvement of upregulation of DEPDC1 (DEP
domain containing 1) in bladder carcinogenesis." Oncogene 26(44): 6448-6455.
Kang, S., Y. B. Kim, et al. (2005). "Polymorphism in the nuclear factor kappa-B binding
promoter region of cyclooxygenase-2 is associated with an increased risk of
bladder cancer." Cancer Lett 217(1): 11-16.
Kao, J., K. Salari, et al. (2009). "Molecular profiling of breast cancer cell lines defines
relevant tumor models and provides a resource for cancer gene discovery." PLoS
One 4(7): e6146.
Karin, M. (1999). "How NF-kappaB is activated: the role of the IkappaB kinase (IKK)
complex." Oncogene 18(49): 6867-6874.
Karin, M. and Y. Ben-Neriah (2000). "Phosphorylation meets ubiquitination: the control
of NF-[kappa]B activity." Annu Rev Immunol 18: 621-663.
Karlseder, J., D. Broccoli, et al. (1999). "p53- and ATM-dependent apoptosis induced by
telomeres lacking TRF2." Science 283(5406): 1321-1325.
Katakura, Y. (2006). "Molecular basis for the cellular senescence program and its
application to anticancer therapy." Biosci Biotechnol Biochem 70(5): 1076-1081.
Katz, M. E. and F. McCormick (1997). "Signal transduction from multiple Ras
effectors." Curr Opin Genet Dev 7(1): 75-79.
Kawakubo, H., J. L. Carey, et al. (2004). "Expression of the NF-kappaB-responsive gene
BTG2 is aberrantly regulated in breast cancer." Oncogene 23(50): 8310-8319.
Keeshan, K., Y. He, et al. (2006). "Tribbles homolog 2 inactivates C/EBPalpha and
causes acute myelogenous leukemia." Cancer Cell 10(5): 401-411.
Kennedy, B. K., D. A. Barbie, et al. (2000). "Nuclear organization of DNA replication in
primary mammalian cells." Genes Dev 14(22): 2855-2868.
Kerzee, J. K. and K. S. Ramos (2000). "Activation of c-Ha-ras by benzo(a)pyrene in
vascular smooth muscle cells involves redox stress and aryl hydrocarbon
receptor." Mol Pharmacol 58(1): 152-158.
Kilbey, A., K. Blyth, et al. (2007). "Runx2 disruption promotes immortalization and
confers resistance to oncogene-induced senescence in primary murine
fibroblasts." Cancer Res 67(23): 11263-11271.
Kim, V. N. (2005). "MicroRNA biogenesis: coordinated cropping and dicing." Nat Rev
Mol Cell Biol 6(5): 376-385.
Kimchi, A., X. F. Wang, et al. (1988). "Absence of TGF-beta receptors and growth
inhibitory responses in retinoblastoma cells." Science 240(4849): 196-199.
Kiyono, T., S. A. Foster, et al. (1998). "Both Rb/p16INK4a inactivation and telomerase
activity are required to immortalize human epithelial cells." Nature 396(6706):
84-88.
Knudsen, E. S. and J. Y. Wang (1996). "Differential regulation of retinoblastoma protein
function by specific Cdk phosphorylation sites." J Biol Chem 271(14): 83138320.
Knudson, A. G., Jr. (1971). "Mutation and cancer: statistical study of retinoblastoma."
Proc Natl Acad Sci U S A 68(4): 820-823.
348
Koli, K., J. Saharinen, et al. (2001). "Novel non-TGF-beta-binding splice variant of
LTBP-4 in human cells and tissues provides means to decrease TGF-beta
deposition." J Cell Sci 114(Pt 15): 2869-2878.
Kriete, A., K. L. Mayo, et al. (2008). "Cell autonomous expression of inflammatory
genes in biologically aged fibroblasts associated with elevated NF-kappaB
activity." Immun Ageing 5: 5.
Krtolica, A. and J. Campisi (2002). "Cancer and aging: a model for the cancer promoting
effects of the aging stroma." Int J Biochem Cell Biol 34(11): 1401-1414.
Krutzfeldt, J., N. Rajewsky, et al. (2005). "Silencing of microRNAs in vivo with
'antagomirs'." Nature 438(7068): 685-689.
Kuhn, A. R., K. Schlauch, et al. (2010). "MicroRNA expression in human airway
smooth muscle cells: role of miR-25 in regulation of airway smooth muscle
phenotype." Am J Respir Cell Mol Biol 42(4): 506-513.
Kuilman, T., C. Michaloglou, et al. (2008). "Oncogene-induced senescence relayed by
an interleukin-dependent inflammatory network." Cell 133(6): 1019-1031.
Kuilman, T. and D. S. Peeper (2009). "Senescence-messaging secretome: SMS-ing
cellular stress." Nat Rev Cancer 9(2): 81-94.
Kumamoto, K., E. A. Spillare, et al. (2008). "Nutlin-3a activates p53 to both downregulate inhibitor of growth 2 and up-regulate mir-34a, mir-34b, and mir-34c
expression, and induce senescence." Cancer Res 68(9): 3193-3203.
Kumar, M. S., J. Lu, et al. (2007). "Impaired microRNA processing enhances cellular
transformation and tumorigenesis." Nat Genet 39(5): 673-677.
Kurochkin, I. V., N. Yonemitsu, et al. (2001). "ALEX1, a novel human armadillo repeat
protein that is expressed differentially in normal tissues and carcinomas."
Biochem Biophys Res Commun 280(1): 340-347.
Kusenda, B., M. Mraz, et al. (2006). "MicroRNA biogenesis, functionality and cancer
relevance." Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub 150(2):
205-215.
LaBaer, J., M. D. Garrett, et al. (1997). "New functional activities for the p21 family of
CDK inhibitors." Genes Dev 11(7): 847-862.
Lachner, M., D. O'Carroll, et al. (2001). "Methylation of histone H3 lysine 9 creates a
binding site for HP1 proteins." Nature 410(6824): 116-120.
Lagos-Quintana, M., R. Rauhut, et al. (2001). "Identification of novel genes coding for
small expressed RNAs." Science 294(5543): 853-858.
Lagos-Quintana, M., R. Rauhut, et al. (2002). "Identification of tissue-specific
microRNAs from mouse." Curr Biol 12(9): 735-739.
Lai, A., B. K. Kennedy, et al. (2001). "RBP1 recruits the mSIN3-histone deacetylase
complex to the pocket of retinoblastoma tumor suppressor family proteins found
in limited discrete regions of the nucleus at growth arrest." Mol Cell Biol 21(8):
2918-2932.
Land, H., A. C. Chen, et al. (1986). "Behavior of myc and ras oncogenes in
transformation of rat embryo fibroblasts." Mol Cell Biol 6(6): 1917-1925.
Land, H., L. F. Parada, et al. (1983). "Tumorigenic conversion of primary embryo
fibroblasts requires at least two cooperating oncogenes." Nature 304(5927): 596602.
Lane, D. P. (1992). "Cancer. p53, guardian of the genome." Nature 358(6381): 15-16.
349
Lane, D. P. and L. V. Crawford (1979). "T antigen is bound to a host protein in SV40transformed cells." Nature 278(5701): 261-263.
Langley, E., M. Pearson, et al. (2002). "Human SIR2 deacetylates p53 and antagonizes
PML/p53-induced cellular senescence." EMBO J 21(10): 2383-2396.
Laoukili, J., M. Alvarez-Fernandez, et al. (2008). "FoxM1 is degraded at mitotic exit in
a Cdh1-dependent manner." Cell Cycle 7(17): 2720-2726.
Laoukili, J., M. R. Kooistra, et al. (2005). "FoxM1 is required for execution of the
mitotic programme and chromosome stability." Nat Cell Biol 7(2): 126-136.
Laoukili, J., M. Stahl, et al. (2007). "FoxM1: at the crossroads of ageing and cancer."
Biochim Biophys Acta 1775(1): 92-102.
Larkin, J. E., B. C. Frank, et al. (2005). "Independence and reproducibility across
microarray platforms." Nat Methods 2(5): 337-344.
Lau, N. C., L. P. Lim, et al. (2001). "An abundant class of tiny RNAs with probable
regulatory roles in Caenorhabditis elegans." Science 294(5543): 858-862.
Lee, E. Y., C. Y. Chang, et al. (1992). "Mice deficient for Rb are nonviable and show
defects in neurogenesis and haematopoiesis." Nature 359(6393): 288-294.
Lee, J. O., A. A. Russo, et al. (1998). "Structure of the retinoblastoma tumoursuppressor pocket domain bound to a peptide from HPV E7." Nature 391(6670):
859-865.
Lee, L. K., B. M. Dunham, et al. (2006). "Cellular dynamics of antisense
oligonucleotides and short interfering RNAs." Ann N Y Acad Sci 1082: 47-51.
Lee, M. L., F. C. Kuo, et al. (2000). "Importance of replication in microarray gene
expression studies: statistical methods and evidence from repetitive cDNA
hybridizations." Proc Natl Acad Sci U S A 97(18): 9834-9839.
Lee, R. C., R. L. Feinbaum, et al. (1993). "The C. elegans heterochronic gene lin-4
encodes small RNAs with antisense complementarity to lin-14." Cell 75(5): 843854.
Lee, W. H., J. Y. Shew, et al. (1987). "The retinoblastoma susceptibility gene encodes a
nuclear phosphoprotein associated with DNA binding activity." Nature
329(6140): 642-645.
Levrero, M., V. De Laurenzi, et al. (2000). "The p53/p63/p73 family of transcription
factors: overlapping and distinct functions." J Cell Sci 113 ( Pt 10): 1661-1670.
Li, Q. and I. M. Verma (2002). "NF-kappaB regulation in the immune system." Nat Rev
Immunol 2(10): 725-734.
Li, R., S. Waga, et al. (1994). "Differential effects by the p21 CDK inhibitor on PCNAdependent DNA replication and repair." Nature 371(6497): 534-537.
Li, X., J. K. Cowell, et al. (2004). "CLCA2 tumour suppressor gene in 1p31 is
epigenetically regulated in breast cancer." Oncogene 23(7): 1474-1480.
Li, X. F., P. J. Yan, et al. (2009). "Downregulation of miR-193b contributes to enhance
urokinase-type plasminogen activator (uPA) expression and tumor progression
and invasion in human breast cancer." Oncogene 28(44): 3937-3948.
Lill, N. L., S. R. Grossman, et al. (1997). "Binding and modulation of p53 by p300/CBP
coactivators." Nature 387(6635): 823-827.
Lim, L. P., N. C. Lau, et al. (2003). "The microRNAs of Caenorhabditis elegans." Genes
Dev 17(8): 991-1008.
350
Lin, A. W., M. Barradas, et al. (1998). "Premature senescence involving p53 and p16 is
activated in response to constitutive MEK/MAPK mitogenic signaling." Genes
Dev 12(19): 3008-3019.
Lin, M. L., J. H. Park, et al. (2007). "Involvement of maternal embryonic leucine zipper
kinase (MELK) in mammary carcinogenesis through interaction with Bcl-G, a
pro-apoptotic member of the Bcl-2 family." Breast Cancer Res 9(1): R17.
Lindstrom, M. S. and K. G. Wiman (2003). "Myc and E2F1 induce p53 through
p14ARF-independent mechanisms in human fibroblasts." Oncogene 22(32):
4993-5005.
Linggi, B., C. Muller-Tidow, et al. (2002). "The t(8;21) fusion protein, AML1 ETO,
specifically represses the transcription of the p14(ARF) tumor suppressor in
acute myeloid leukemia." Nat Med 8(7): 743-750.
Linzer, D. I. and A. J. Levine (1979). "Characterization of a 54K dalton cellular SV40
tumor antigen present in SV40-transformed cells and uninfected embryonal
carcinoma cells." Cell 17(1): 43-52.
Lipinski, M. M. and T. Jacks (1999). "The retinoblastoma gene family in differentiation
and development." Oncogene 18(55): 7873-7882.
Liu, C. G., G. A. Calin, et al. (2004). "An oligonucleotide microchip for genome-wide
microRNA profiling in human and mouse tissues." Proc Natl Acad Sci U S A
101(26): 9740-9744.
Liu, L., N. J. Lassam, et al. (1995). "Germline p16INK4A mutation and protein
dysfunction in a family with inherited melanoma." Oncogene 11(2): 405-412.
Lockyer, P. J., S. Kupzig, et al. (2001). "CAPRI regulates Ca(2+)-dependent inactivation
of the Ras-MAPK pathway." Curr Biol 11(12): 981-986.
Lohrum, M. A. and K. H. Vousden (2000). "Regulation and function of the p53-related
proteins: same family, different rules." Trends Cell Biol 10(5): 197-202.
Lowe, S. W. and H. E. Ruley (1993). "Stabilization of the p53 tumor suppressor is
induced by adenovirus 5 E1A and accompanies apoptosis." Genes Dev 7(4): 535545.
Lowe, S. W. and C. J. Sherr (2003). "Tumor suppression by Ink4a-Arf: progress and
puzzles." Curr Opin Genet Dev 13(1): 77-83.
Lu, T. and T. Finkel (2008). "Free radicals and senescence." Exp Cell Res 314(9): 19181922.
Lu, W., R. Pochampally, et al. (2000). "Nuclear exclusion of p53 in a subset of tumors
requires MDM2 function." Oncogene 19(2): 232-240.
Ludlow, J. W., J. A. DeCaprio, et al. (1989). "SV40 large T antigen binds preferentially
to an underphosphorylated member of the retinoblastoma susceptibility gene
product family." Cell 56(1): 57-65.
Ludlow, J. W., J. Shon, et al. (1990). "The retinoblastoma susceptibility gene product
undergoes cell cycle-dependent dephosphorylation and binding to and release
from SV40 large T." Cell 60(3): 387-396.
Lukas, J., D. Parry, et al. (1995). "Retinoblastoma-protein-dependent cell-cycle
inhibition by the tumour suppressor p16." Nature 375(6531): 503-506.
Luo, R. X., A. A. Postigo, et al. (1998). "Rb interacts with histone deacetylase to repress
transcription." Cell 92(4): 463-473.
Ma, L. and R. A. Weinberg (2008). "MicroRNAs in malignant progression." Cell Cycle
7(5): 570-572.
351
Ma, R. Y., T. H. Tong, et al. (2005). "Raf/MEK/MAPK signaling stimulates the nuclear
translocation and transactivating activity of FOXM1c." J Cell Sci 118(Pt 4): 795806.
Macip, S., M. Igarashi, et al. (2003). "Influence of induced reactive oxygen species in
p53-mediated cell fate decisions." Mol Cell Biol 23(23): 8576-8585.
Macip, S., M. Igarashi, et al. (2002). "Inhibition of p21-mediated ROS accumulation can
rescue p21-induced senescence." EMBO J 21(9): 2180-2188.
Maeda, T., R. M. Hobbs, et al. (2005). "Role of the proto-oncogene Pokemon in cellular
transformation and ARF repression." Nature 433(7023): 278-285.
Maier, B., W. Gluba, et al. (2004). "Modulation of mammalian life span by the short
isoform of p53." Genes Dev 18(3): 306-319.
Maione, R., G. M. Fimia, et al. (1994). "Retinoblastoma antioncogene is involved in the
inhibition of myogenesis by polyomavirus large T antigen." Cell Growth Differ
5(2): 231-237.
Mantovani, F. and L. Banks (2001). "The human papillomavirus E6 protein and its
contribution to malignant progression." Oncogene 20(54): 7874-7887.
Mao, L., A. Merlo, et al. (1995). "A novel p16INK4A transcript." Cancer Res 55(14):
2995-2997.
Marasa, B. S., S. Srikantan, et al. (2009). "Increased MKK4 abundance with replicative
senescence is linked to the joint reduction of multiple microRNAs." Sci Signal
2(94): ra69.
Marie, S. K., O. K. Okamoto, et al. (2008). "Maternal embryonic leucine zipper kinase
transcript abundance correlates with malignancy grade in human astrocytomas."
Int J Cancer 122(4): 807-815.
Martens, U. M., E. A. Chavez, et al. (2000). "Accumulation of short telomeres in human
fibroblasts prior to replicative senescence." Exp Cell Res 256(1): 291-299.
Martin, D. W., M. A. Subler, et al. (1993). "p53 and SV40 T antigen bind to the same
region overlapping the conserved domain of the TATA-binding protein."
Biochem Biophys Res Commun 195(1): 428-434.
Martin, K. J., E. Graner, et al. (2001). "High-sensitivity array analysis of gene
expression for the early detection of disseminated breast tumor cells in peripheral
blood." Proc Natl Acad Sci U S A 98(5): 2646-2651.
Martinez, I., A. S. Gardiner, et al. (2008). "Human papillomavirus type 16 reduces the
expression of microRNA-218 in cervical carcinoma cells." Oncogene 27(18):
2575-2582.
Marz, P., A. Probst, et al. (2004). "Ataxin-10, the spinocerebellar ataxia type 10
neurodegenerative disorder protein, is essential for survival of cerebellar
neurons." J Biol Chem 279(34): 35542-35550.
Maser, R. S. and R. A. DePinho (2002). "Connecting chromosomes, crisis, and cancer."
Science 297(5581): 565-569.
Massague, J. (1990). "Transforming growth factor-alpha. A model for membraneanchored growth factors." J Biol Chem 265(35): 21393-21396.
Massimi, P. and L. Banks (2000). "Differential phosphorylation of the HPV-16 E7
oncoprotein during the cell cycle." Virology 276(2): 388-394.
Matsuda, A., Y. Suzuki, et al. (2003). "Large-scale identification and characterization of
human genes that activate NF-kappaB and MAPK signaling pathways."
Oncogene 22(21): 3307-3318.
352
Matsumoto, N., N. Yoneda-Kato, et al. (2000). "Elevated MLF1 expression correlates
with malignant progression from myelodysplastic syndrome." Leukemia 14(10):
1757-1765.
Mattes, J., A. Collison, et al. (2008). "Emerging role of microRNAs in disease
pathogenesis and strategies for therapeutic modulation." Curr Opin Mol Ther
10(2): 150-157.
Mayol, X., X. Grana, et al. (1993). "Cloning of a new member of the retinoblastoma
gene family (pRb2) which binds to the E1A transforming domain." Oncogene
8(9): 2561-2566.
Mayr, C., M. T. Hemann, et al. (2007). "Disrupting the pairing between let-7 and Hmga2
enhances oncogenic transformation." Science 315(5818): 1576-1579.
McConnell, B. B., M. Starborg, et al. (1998). "Inhibitors of cyclin-dependent kinases
induce features of replicative senescence in early passage human diploid
fibroblasts." Curr Biol 8(6): 351-354.
McMahon, M. and D. Woods (2001). "Regulation of the p53 pathway by Ras, the plot
thickens." Biochim Biophys Acta 1471(2): M63-71.
McPherron, A. C. and S. J. Lee (1997). "Double muscling in cattle due to mutations in
the myostatin gene." Proc Natl Acad Sci U S A 94(23): 12457-12461.
Mertens-Talcott, S. U., S. Chintharlapalli, et al. (2007). "The oncogenic microRNA-27a
targets genes that regulate specificity protein transcription factors and the G2-M
checkpoint in MDA-MB-231 breast cancer cells." Cancer Res 67(22): 1100111011.
Michael, D. and M. Oren (2002). "The p53 and Mdm2 families in cancer." Curr Opin
Genet Dev 12(1): 53-59.
Michaloglou, C., L. C. Vredeveld, et al. (2005). "BRAFE600-associated senescence-like
cell cycle arrest of human naevi." Nature 436(7051): 720-724.
Millis, A. J., H. M. McCue, et al. (1992). "Metalloproteinase and TIMP-1 gene
expression during replicative senescence." Exp Gerontol 27(4): 425-428.
Missero, C., F. Di Cunto, et al. (1996). "The absence of p21Cip1/WAF1 alters
keratinocyte growth and differentiation and promotes ras-tumor progression."
Genes Dev 10(23): 3065-3075.
Mitchell, P. J., C. Wang, et al. (1987). "Positive and negative regulation of transcription
in vitro: enhancer-binding protein AP-2 is inhibited by SV40 T antigen." Cell
50(6): 847-861.
Mizoi, T., H. Ohtani, et al. (1993). "Immunoelectron microscopic localization of
transforming growth factor beta 1 and latent transforming growth factor beta 1
binding protein in human gastrointestinal carcinomas: qualitative difference
between cancer cells and stromal cells." Cancer Res 53(1): 183-190.
Mohr, S., G. D. Leikauf, et al. (2002). "Microarrays as cancer keys: an array of
possibilities." J Clin Oncol 20(14): 3165-3175.
Momand, J., D. Jung, et al. (1998). "The MDM2 gene amplification database." Nucleic
Acids Res 26(15): 3453-3459.
Montes de Oca Luna, R., D. S. Wagner, et al. (1995). "Rescue of early embryonic
lethality in mdm2-deficient mice by deletion of p53." Nature 378(6553): 203206.
Mooi, W. J. and D. S. Peeper (2006). "Oncogene-induced cell senescence--halting on the
road to cancer." N Engl J Med 355(10): 1037-1046.
353
Morales, C. P., S. E. Holt, et al. (1999). "Absence of cancer-associated changes in
human fibroblasts immortalized with telomerase." Nat Genet 21(1): 115-118.
Moran, E. (1988). "A region of SV40 large T antigen can substitute for a transforming
domain of the adenovirus E1A products." Nature 334(6178): 168-170.
Moran, E. (1993). "Interaction of adenoviral proteins with pRB and p53." FASEB J
7(10): 880-885.
Morris, E. J. and N. J. Dyson (2001). "Retinoblastoma protein partners." Adv Cancer
Res 82: 1-54.
Motoda, L., M. Osato, et al. (2007). "Runx1 protects hematopoietic stem/progenitor cells
from oncogenic insult." Stem Cells 25(12): 2976-2986.
Moyzis, R. K., J. M. Buckingham, et al. (1988). "A highly conserved repetitive DNA
sequence, (TTAGGG)n, present at the telomeres of human chromosomes." Proc
Natl Acad Sci U S A 85(18): 6622-6626.
Mraz, M., S. Pospisilova, et al. (2009). "MicroRNAs in chronic lymphocytic leukemia
pathogenesis and disease subtypes." Leuk Lymphoma 50(3): 506-509.
Munger, K., B. A. Werness, et al. (1989). "Complex formation of human papillomavirus
E7 proteins with the retinoblastoma tumor suppressor gene product." EMBO J
8(13): 4099-4105.
Munro, J., F. J. Stott, et al. (1999). "Role of the alternative INK4A proteins in human
keratinocyte senescence: evidence for the specific inactivation of p16INK4A
upon immortalization." Cancer Res 59(11): 2516-2521.
Muntoni, A. and R. R. Reddel (2005). "The first molecular details of ALT in human
tumor cells." Hum Mol Genet 14 Spec No. 2: R191-196.
Murphree, A. L. and W. F. Benedict (1984). "Retinoblastoma: clues to human
oncogenesis." Science 223(4640): 1028-1033.
Murphy, M., J. Ahn, et al. (1999). "Transcriptional repression by wild-type p53 utilizes
histone deacetylases, mediated by interaction with mSin3a." Genes Dev 13(19):
2490-2501.
Myatt, S. S. and E. W. Lam (2007). "The emerging roles of forkhead box (Fox) proteins
in cancer." Nat Rev Cancer 7(11): 847-859.
Mymryk, J. S. and S. T. Bayley (1994). "Multiple pathways for activation of E2A
expression in human KB cells by the 243R E1A protein of adenovirus type 5."
Virus Res 33(1): 89-97.
Nabel, G. J. and I. M. Verma (1993). "Proposed NF-kappa B/I kappa B family
nomenclature." Genes Dev 7(11): 2063.
Nagai, H., M. Negrini, et al. (1995). "Detection and cloning of a common region of loss
of heterozygosity at chromosome 1p in breast cancer." Cancer Res 55(8): 17521757.
Naiki, T., E. Saijou, et al. (2007). "TRB2, a mouse Tribbles ortholog, suppresses
adipocyte differentiation by inhibiting AKT and C/EBPbeta." J Biol Chem
282(33): 24075-24082.
Nakano, I., M. Masterman-Smith, et al. (2008). "Maternal embryonic leucine zipper
kinase is a key regulator of the proliferation of malignant brain tumors, including
brain tumor stem cells." J Neurosci Res 86(1): 48-60.
Narayanan, B. A., N. K. Narayanan, et al. (2006). "RNA interference-mediated
cyclooxygenase-2 inhibition prevents prostate cancer cell growth and induces
354
differentiation: modulation of neuronal protein synaptophysin, cyclin D1, and
androgen receptor." Mol Cancer Ther 5(5): 1117-1125.
Narita, M. and S. W. Lowe (2005). "Senescence comes of age." Nat Med 11(9): 920922.
Narita, M., S. Nunez, et al. (2003). "Rb-mediated heterochromatin formation and
silencing of E2F target genes during cellular senescence." Cell 113(6): 703-716.
Nelson, D. A., N. A. Krucher, et al. (1997). "High molecular weight protein phosphatase
type 1 dephosphorylates the retinoblastoma protein." J Biol Chem 272(7): 45284535.
Nevels, M., S. Rubenwolf, et al. (1997). "The adenovirus E4orf6 protein can promote
E1A/E1B-induced focus formation by interfering with p53 tumor suppressor
function." Proc Natl Acad Sci U S A 94(4): 1206-1211.
Nevins, J. R. (2001). "The Rb/E2F pathway and cancer." Hum Mol Genet 10(7): 699703.
Nielsen, S. J., R. Schneider, et al. (2001). "Rb targets histone H3 methylation and HP1
to promoters." Nature 412(6846): 561-565.
Nip, J., D. K. Strom, et al. (2001). "E2F-1 induces the stabilization of p53 but blocks
p53-mediated transactivation." Oncogene 20(8): 910-920.
Noda, A., Y. Ning, et al. (1994). "Cloning of senescent cell-derived inhibitors of DNA
synthesis using an expression screen." Exp Cell Res 211(1): 90-98.
O'Donnell, K. A., E. A. Wentzel, et al. (2005). "c-Myc-regulated microRNAs modulate
E2F1 expression." Nature 435(7043): 839-843.
O'Hare, M. J., J. Bond, et al. (2001). "Conditional immortalization of freshly isolated
human mammary fibroblasts and endothelial cells." Proc Natl Acad Sci U S A
98(2): 646-651.
Ogami, M., Y. Ikura, et al. (2004). "Telomere shortening in human coronary artery
diseases." Arterioscler Thromb Vasc Biol 24(3): 546-550.
Oh, S. T., S. Kyo, et al. (2001). "Telomerase activation by human papillomavirus type
16 E6 protein: induction of human telomerase reverse transcriptase expression
through Myc and GC-rich Sp1 binding sites." J Virol 75(12): 5559-5566.
Ohtani, N., Z. Zebedee, et al. (2001). "Opposing effects of Ets and Id proteins on
p16INK4a expression during cellular senescence." Nature 409(6823): 10671070.
Oklu, R. and R. Hesketh (2000). "The latent transforming growth factor beta binding
protein (LTBP) family." Biochem J 352 Pt 3: 601-610.
Oren, M. and A. J. Levine (1981). "Immunoselection of simian virus 40 large T antigen
messenger rnas from transformed cells." Virology 113(2): 790-793.
Orjalo, A. V., D. Bhaumik, et al. (2009). "Cell surface-bound IL-1alpha is an upstream
regulator of the senescence-associated IL-6/IL-8 cytokine network." Proc Natl
Acad Sci U S A 106(40): 17031-17036.
Ortega, S., M. Malumbres, et al. (2002). "Cyclin D-dependent kinases, INK4 inhibitors
and cancer." Biochim Biophys Acta 1602(1): 73-87.
Osada, M., M. Ohba, et al. (1998). "Cloning and functional analysis of human p51,
which structurally and functionally resembles p53." Nat Med 4(7): 839-843.
Ossovskaya, V. S., I. A. Mazo, et al. (1996). "Use of genetic suppressor elements to
dissect distinct biological effects of separate p53 domains." Proc Natl Acad Sci
U S A 93(19): 10309-10314.
355
Otterson, G. A., R. A. Kratzke, et al. (1994). "Absence of p16INK4 protein is restricted
to the subset of lung cancer lines that retains wildtype RB." Oncogene 9(11):
3375-3378.
Paddison, P. J. (2008). "RNA interference in mammalian cell systems." Curr Top
Microbiol Immunol 320: 1-19.
Paddison, P. J., J. M. Silva, et al. (2004). "A resource for large-scale RNA-interferencebased screens in mammals." Nature 428(6981): 427-431.
Palmero, I., C. Pantoja, et al. (1998). "p19ARF links the tumour suppressor p53 to Ras."
Nature 395(6698): 125-126.
Panossian, L. A., V. R. Porter, et al. (2003). "Telomere shortening in T cells correlates
with Alzheimer's disease status." Neurobiol Aging 24(1): 77-84.
Pantoja, C. and M. Serrano (1999). "Murine fibroblasts lacking p21 undergo senescence
and are resistant to transformation by oncogenic Ras." Oncogene 18(35): 49744982.
Parada, L. F., C. J. Tabin, et al. (1982). "Human EJ bladder carcinoma oncogene is
homologue of Harvey sarcoma virus ras gene." Nature 297(5866): 474-478.
Parisi, T., A. Pollice, et al. (2002). "Transcriptional regulation of the human tumor
suppressor p14(ARF) by E2F1, E2F2, E2F3, and Sp1-like factors." Biochem
Biophys Res Commun 291(5): 1138-1145.
Park, H. J., Z. Wang, et al. (2008). "An N-terminal inhibitory domain modulates activity
of FoxM1 during cell cycle." Oncogene 27(12): 1696-1704.
Park, I. K., S. J. Morrison, et al. (2004). "Bmi1, stem cells, and senescence regulation." J
Clin Invest 113(2): 175-179.
Parrinello, S., E. Samper, et al. (2003). "Oxygen sensitivity severely limits the
replicative lifespan of murine fibroblasts." Nat Cell Biol 5(8): 741-747.
Parsons, D. W., S. Jones, et al. (2008). "An integrated genomic analysis of human
glioblastoma multiforme." Science 321(5897): 1807-1812.
Pascal, T., F. Debacq-Chainiaux, et al. (2005). "Comparison of replicative senescence
and stress-induced premature senescence combining differential display and lowdensity DNA arrays." FEBS Lett 579(17): 3651-3659.
Passos, J. F., G. Nelson, et al. (2010). "Feedback between p21 and reactive oxygen
production is necessary for cell senescence." Mol Syst Biol 6: 347.
Passos, J. F., G. Saretzki, et al. (2007). "Mitochondrial dysfunction accounts for the
stochastic heterogeneity in telomere-dependent senescence." PLoS Biol 5(5):
e110.
Passos, J. F., G. Saretzki, et al. (2007). "DNA damage in telomeres and mitochondria
during cellular senescence: is there a connection?" Nucleic Acids Res 35(22):
7505-7513.
Pavel, S., N. P. Smit, et al. (2003). "Homozygous germline mutation of CDKN2A/p16
and glucose-6-phosphate dehydrogenase deficiency in a multiple melanoma
case." Melanoma Res 13(2): 171-178.
Pearson, M., R. Carbone, et al. (2000). "PML regulates p53 acetylation and premature
senescence induced by oncogenic Ras." Nature 406(6792): 207-210.
Pedeux, R., S. Sengupta, et al. (2005). "ING2 regulates the onset of replicative
senescence by induction of p300-dependent p53 acetylation." Mol Cell Biol
25(15): 6639-6648.
356
Penttinen, C., J. Saharinen, et al. (2002). "Secretion of human latent TGF-beta-binding
protein-3 (LTBP-3) is dependent on co-expression of TGF-beta." J Cell Sci
115(Pt 17): 3457-3468.
Penzo, M., P. E. Massa, et al. (2009). "Sustained NF-kappaB activation produces a
short-term cell proliferation block in conjunction with repressing effectors of cell
cycle progression controlled by E2F or FoxM1." J Cell Physiol 218(1): 215-227.
Perricaudet, M., J. M. le Moullec, et al. (1980). "Structure of two adenovirus type 12
transforming polypeptides and their evolutionary implications." Nature
288(5787): 174-176.
Perry, M. M., A. E. Williams, et al. (2009). "Divergent intracellular pathways regulate
interleukin-1beta-induced miR-146a and miR-146b expression and chemokine
release in human alveolar epithelial cells." FEBS Lett 583(20): 3349-3355.
Phillips, A. C. and K. H. Vousden (2001). "E2F-1 induced apoptosis." Apoptosis 6(3):
173-182.
Piedimonte, G., A. F. Borghetti, et al. (1982). "Effect of cell density on growth rate and
amino acid transport in simian virus 40-transformed 3T3 cells." Cancer Res
42(11): 4690-4693.
Place, R. F., L. C. Li, et al. (2008). "MicroRNA-373 induces expression of genes with
complementary promoter sequences." Proc Natl Acad Sci U S A 105(5): 16081613.
Poliseno, L., L. Pitto, et al. (2008). "The proto-oncogene LRF is under posttranscriptional control of MiR-20a: implications for senescence." PLoS One 3(7):
e2542.
Polyak, K., Y. Xia, et al. (1997). "A model for p53-induced apoptosis." Nature
389(6648): 300-305.
Pomerantz, J., N. Schreiber-Agus, et al. (1998). "The Ink4a tumor suppressor gene
product, p19Arf, interacts with MDM2 and neutralizes MDM2's inhibition of
p53." Cell 92(6): 713-723.
Powell, A. J., A. J. Darmon, et al. (1999). "Different functions are required for initiation
and maintenance of immortalization of rat embryo fibroblasts by SV40 large T
antigen." Oncogene 18(51): 7343-7350.
Poy, M. N., L. Eliasson, et al. (2004). "A pancreatic islet-specific microRNA regulates
insulin secretion." Nature 432(7014): 226-230.
Prives, C. (1998). "Signaling to p53: breaking the MDM2-p53 circuit." Cell 95(1): 5-8.
Qi, J., J. Y. Yu, et al. (2009). "microRNAs regulate human embryonic stem cell
division." Cell Cycle 8(22): 3729-3741.
Quartin, R. S., C. N. Cole, et al. (1994). "The amino-terminal functions of the simian
virus 40 large T antigen are required to overcome wild-type p53-mediated
growth arrest of cells." J Virol 68(3): 1334-1341.
Quelle, D. E., F. Zindy, et al. (1995). "Alternative reading frames of the INK4a tumor
suppressor gene encode two unrelated proteins capable of inducing cell cycle
arrest." Cell 83(6): 993-1000.
Rai, R., A. Phadnis, et al. (2008). "Differential regulation of centrosome integrity by
DNA damage response proteins." Cell Cycle 7(14): 2225-2233.
Ramaswamy, S. and C. M. Perou (2003). "DNA microarrays in breast cancer: the
promise of personalised medicine." Lancet 361(9369): 1576-1577.
357
Ramsey, M. R. and N. E. Sharpless (2006). "ROS as a tumour suppressor?" Nat Cell
Biol 8(11): 1213-1215.
Rao, L., M. Debbas, et al. (1992). "The adenovirus E1A proteins induce apoptosis,
which is inhibited by the E1B 19-kDa and Bcl-2 proteins." Proc Natl Acad Sci U
S A 89(16): 7742-7746.
Rapp, L. and J. J. Chen (1998). "The papillomavirus E6 proteins." Biochim Biophys
Acta 1378(1): F1-19.
Rauhala, H. E., S. E. Jalava, et al. (2010). "miR-193b is an epigenetically regulated
putative tumor suppressor in prostate cancer." Int J Cancer 127(6): 1363-1372.
Rayman, J. B., Y. Takahashi, et al. (2002). "E2F mediates cell cycle-dependent
transcriptional repression in vivo by recruitment of an HDAC1/mSin3B
corepressor complex." Genes Dev 16(8): 933-947.
Rebbaa, A., X. Zheng, et al. (2003). "Caspase inhibition switches doxorubicin-induced
apoptosis to senescence." Oncogene 22(18): 2805-2811.
Reinhart, B. J., E. G. Weinstein, et al. (2002). "MicroRNAs in plants." Genes Dev
16(13): 1616-1626.
Reynolds, A., D. Leake, et al. (2004). "Rational siRNA design for RNA interference."
Nat Biotechnol 22(3): 326-330.
Rheinwald, J. G., W. C. Hahn, et al. (2002). "A two-stage, p16(INK4A)- and p53dependent keratinocyte senescence mechanism that limits replicative potential
independent of telomere status." Mol Cell Biol 22(14): 5157-5172.
Rhoades, M. W., B. J. Reinhart, et al. (2002). "Prediction of plant microRNA targets."
Cell 110(4): 513-520.
Rhodes, D. R., J. Yu, et al. (2004). "Large-scale meta-analysis of cancer microarray data
identifies common transcriptional profiles of neoplastic transformation and
progression." Proc Natl Acad Sci U S A 101(25): 9309-9314.
Rice, P. W. and C. N. Cole (1993). "Efficient transcriptional activation of many simple
modular promoters by simian virus 40 large T antigen." J Virol 67(11): 66896697.
Roberts, J. M. (1999). "Evolving ideas about cyclins." Cell 98(2): 129-132.
Rodier, F., J. P. Coppe, et al. (2009). "Persistent DNA damage signalling triggers
senescence-associated inflammatory cytokine secretion." Nat Cell Biol 11(8):
973-979.
Rodriguez, A., S. Griffiths-Jones, et al. (2004). "Identification of mammalian microRNA
host genes and transcription units." Genome Res 14(10A): 1902-1910.
Rogoff, H. A., M. T. Pickering, et al. (2002). "E2F1 induces phosphorylation of p53 that
is coincident with p53 accumulation and apoptosis." Mol Cell Biol 22(15): 53085318.
Rowland, B. D., S. G. Denissov, et al. (2002). "E2F transcriptional repressor complexes
are critical downstream targets of p19(ARF)/p53-induced proliferative arrest."
Cancer Cell 2(1): 55-65.
Rushton, J. J., D. Jiang, et al. (1997). "Simian virus 40 T antigen can regulate p53mediated transcription independent of binding p53." J Virol 71(7): 5620-5623.
Saffer, J. D., S. P. Jackson, et al. (1990). "SV40 stimulates expression of the transacting
factor Sp1 at the mRNA level." Genes Dev 4(4): 659-666.
358
Salminen, A., J. Ojala, et al. (2008). "Interaction of aging-associated signaling cascades:
inhibition of NF-kappaB signaling by longevity factors FoxOs and SIRT1." Cell
Mol Life Sci 65(7-8): 1049-1058.
Sang, L., H. A. Coller, et al. (2008). "Control of the reversibility of cellular quiescence
by the transcriptional repressor HES1." Science 321(5892): 1095-1100.
Saretzki, G. (2003). "Telomerase inhibition as cancer therapy." Cancer Lett 194(2): 209219.
Sato, F., N. Harpaz, et al. (2002). "Hypermethylation of the p14(ARF) gene in ulcerative
colitis-associated colorectal carcinogenesis." Cancer Res 62(4): 1148-1151.
Scheffner, M., J. M. Huibregtse, et al. (1993). "The HPV-16 E6 and E6-AP complex
functions as a ubiquitin-protein ligase in the ubiquitination of p53." Cell 75(3):
495-505.
Scheffner, M., R. Knippers, et al. (1989). "RNA unwinding activity of SV40 large T
antigen." Cell 57(6): 955-963.
Scheffner, M., K. Munger, et al. (1992). "Targeted degradation of the retinoblastoma
protein by human papillomavirus E7-E6 fusion proteins." EMBO J 11(7): 24252431.
Scheidereit, C. (2006). "IkappaB kinase complexes: gateways to NF-kappaB activation
and transcription." Oncogene 25(51): 6685-6705.
Schmale, H. and C. Bamberger (1997). "A novel protein with strong homology to the
tumor suppressor p53." Oncogene 15(11): 1363-1367.
Schmitt, C. A., J. S. Fridman, et al. (2002). "A senescence program controlled by p53
and p16INK4a contributes to the outcome of cancer therapy." Cell 109(3): 335346.
Schwarz, E., U. K. Freese, et al. (1985). "Structure and transcription of human
papillomavirus sequences in cervical carcinoma cells." Nature 314(6006): 111114.
Schwarze, S. R., Y. Shi, et al. (2001). "Role of cyclin-dependent kinase inhibitors in the
growth arrest at senescence in human prostate epithelial and uroepithelial cells."
Oncogene 20(57): 8184-8192.
Sebastian, T., R. Malik, et al. (2005). "C/EBPbeta cooperates with RB:E2F to implement
Ras(V12)-induced cellular senescence." EMBO J 24(18): 3301-3312.
Seger, Y. R., M. Garcia-Cao, et al. (2002). "Transformation of normal human cells in the
absence of telomerase activation." Cancer Cell 2(5): 401-413.
Sellers, W. R. and W. G. Kaelin, Jr. (1997). "Role of the retinoblastoma protein in the
pathogenesis of human cancer." J Clin Oncol 15(11): 3301-3312.
Semizarov, D., L. Frost, et al. (2003). "Specificity of short interfering RNA determined
through gene expression signatures." Proc Natl Acad Sci U S A 100(11): 63476352.
Sen, R. and D. Baltimore (1986). "Inducibility of kappa immunoglobulin enhancerbinding protein Nf-kappa B by a posttranslational mechanism." Cell 47(6): 921928.
Senftleben, U., Y. Cao, et al. (2001). "Activation by IKKalpha of a second, evolutionary
conserved, NF-kappa B signaling pathway." Science 293(5534): 1495-1499.
Serrano, M. (1997). "The tumor suppressor protein p16INK4a." Exp Cell Res 237(1): 713.
359
Serrano, M. and M. A. Blasco (2001). "Putting the stress on senescence." Curr Opin Cell
Biol 13(6): 748-753.
Serrano, M., G. J. Hannon, et al. (1993). "A new regulatory motif in cell-cycle control
causing specific inhibition of cyclin D/CDK4." Nature 366(6456): 704-707.
Serrano, M., H. Lee, et al. (1996). "Role of the INK4a locus in tumor suppression and
cell mortality." Cell 85(1): 27-37.
Serrano, M., A. W. Lin, et al. (1997). "Oncogenic ras provokes premature cell
senescence associated with accumulation of p53 and p16INK4a." Cell 88(5):
593-602.
Shapiro, G. I., J. E. Park, et al. (1995). "Multiple mechanisms of p16INK4A inactivation
in non-small cell lung cancer cell lines." Cancer Res 55(24): 6200-6209.
Sharpless, N. E. and R. A. DePinho (2005). "Cancer: crime and punishment." Nature
436(7051): 636-637.
Shay, J. W., O. M. Pereira-Smith, et al. (1991). "A role for both RB and p53 in the
regulation of human cellular senescence." Exp Cell Res 196(1): 33-39.
Shelton, D. N., E. Chang, et al. (1999). "Microarray analysis of replicative senescence."
Curr Biol 9(17): 939-945.
Sherr, C. J. (1996). "Cancer cell cycles." Science 274(5293): 1672-1677.
Sherr, C. J. (2000). "Cell cycle control and cancer." Harvey Lect 96: 73-92.
Sherr, C. J. (2001). "The INK4a/ARF network in tumour suppression." Nat Rev Mol
Cell Biol 2(10): 731-737.
Sherr, C. J. and R. A. DePinho (2000). "Cellular senescence: mitotic clock or culture
shock?" Cell 102(4): 407-410.
Sherr, C. J. and J. M. Roberts (1999). "CDK inhibitors: positive and negative regulators
of G1-phase progression." Genes Dev 13(12): 1501-1512.
Sherr, C. J. and J. D. Weber (2000). "The ARF/p53 pathway." Curr Opin Genet Dev
10(1): 94-99.
Shivji, M. K., S. J. Grey, et al. (1994). "Cip1 inhibits DNA replication but not PCNAdependent nucleotide excision-repair." Curr Biol 4(12): 1062-1068.
Shohat, G., G. Shani, et al. (2002). "The DAP-kinase family of proteins: study of a novel
group of calcium-regulated death-promoting kinases." Biochim Biophys Acta
1600(1-2): 45-50.
Shvarts, A., T. R. Brummelkamp, et al. (2002). "A senescence rescue screen identifies
BCL6 as an inhibitor of anti-proliferative p19(ARF)-p53 signaling." Genes Dev
16(6): 681-686.
Silberstein, G. B. and C. W. Daniel (1987). "Reversible inhibition of mammary gland
growth by transforming growth factor-beta." Science 237(4812): 291-293.
Sionov, R. V. and Y. Haupt (1999). "The cellular response to p53: the decision between
life and death." Oncogene 18(45): 6145-6157.
Slack, R. S., I. S. Skerjanc, et al. (1995). "Cells differentiating into neuroectoderm
undergo apoptosis in the absence of functional retinoblastoma family proteins." J
Cell Biol 129(3): 779-788.
Smogorzewska, A. and T. de Lange (2002). "Different telomere damage signaling
pathways in human and mouse cells." EMBO J 21(16): 4338-4348.
Sossey-Alaoui, K., E. Kitamura, et al. (2001). "Fine mapping of the PTGFR gene to
1p31 region and mutation analysis in human breast cancer." Int J Mol Med 7(5):
543-546.
360
Soule, H. R. and J. S. Butel (1979). "Subcellular Localization of simian virus 40 large
tumor antigen." J Virol 30(2): 523-532.
Soussi, T. and P. May (1996). "Structural aspects of the p53 protein in relation to gene
evolution: a second look." J Mol Biol 260(5): 623-637.
Spurgers, K. B., C. M. Sharkey, et al. (2008). "Oligonucleotide antiviral therapeutics:
antisense and RNA interference for highly pathogenic RNA viruses." Antiviral
Res 78(1): 26-36.
Stansel, R. M., T. de Lange, et al. (2001). "T-loop assembly in vitro involves binding of
TRF2 near the 3' telomeric overhang." EMBO J 20(19): 5532-5540.
Stein, G. H., L. F. Drullinger, et al. (1999). "Differential roles for cyclin-dependent
kinase inhibitors p21 and p16 in the mechanisms of senescence and
differentiation in human fibroblasts." Mol Cell Biol 19(3): 2109-2117.
Stepkowski, S. M., X. Qu, et al. (2000). "Inhibition of C-raf expression by antisense
oligonucleotides extends heart allograft survival in rats." Transplantation 70(4):
656-661.
Stewart, S. A. and R. A. Weinberg (2006). "Telomeres: cancer to human aging." Annu
Rev Cell Dev Biol 22: 531-557.
Stros, M., T. Ozaki, et al. (2002). "HMGB1 and HMGB2 cell-specifically down-regulate
the p53- and p73-dependent sequence-specific transactivation from the human
Bax gene promoter." J Biol Chem 277(9): 7157-7164.
Su, G., T. Roberts, et al. (1999). "TTC4, a novel human gene containing the
tetratricopeptide repeat and mapping to the region of chromosome 1p31 that is
frequently deleted in sporadic breast cancer." Genomics 55(2): 157-163.
Sun, P., N. Yoshizuka, et al. (2007). "PRAK is essential for ras-induced senescence and
tumor suppression." Cell 128(2): 295-308.
Sun, W., K. Zhang, et al. (2004). "Identification of differentially expressed genes in
human lung squamous cell carcinoma using suppression subtractive
hybridization." Cancer Lett 212(1): 83-93.
Suzuki, M., S. Okuyama, et al. (1998). "A novel E2F binding protein with Myc-type
HLH motif stimulates E2F-dependent transcription by forming a heterodimer."
Oncogene 17(7): 853-865.
Taganov, K. D., M. P. Boldin, et al. (2006). "NF-kappaB-dependent induction of
microRNA miR-146, an inhibitor targeted to signaling proteins of innate immune
responses." Proc Natl Acad Sci U S A 103(33): 12481-12486.
Takahashi, Y., J. B. Rayman, et al. (2000). "Analysis of promoter binding by the E2F
and pRB families in vivo: distinct E2F proteins mediate activation and
repression." Genes Dev 14(7): 804-816.
Tam, W. and J. E. Dahlberg (2006). "miR-155/BIC as an oncogenic microRNA." Genes
Chromosomes Cancer 45(2): 211-212.
Tang, J., G. M. Gordon, et al. (2002). "The helix-loop-helix protein id-1 delays onset of
replicative senescence in human endothelial cells." Lab Invest 82(8): 1073-1079.
Tanzer, A. and P. F. Stadler (2004). "Molecular evolution of a microRNA cluster." J
Mol Biol 339(2): 327-335.
Tao, W. and A. J. Levine (1999). "Nucleocytoplasmic shuttling of oncoprotein Hdm2 is
required for Hdm2-mediated degradation of p53." Proc Natl Acad Sci U S A
96(6): 3077-3080.
361
Tao, W. and A. J. Levine (1999). "P19(ARF) stabilizes p53 by blocking nucleocytoplasmic shuttling of Mdm2." Proc Natl Acad Sci U S A 96(12): 6937-6941.
Taubert, S., C. Gorrini, et al. (2004). "E2F-dependent histone acetylation and
recruitment of the Tip60 acetyltransferase complex to chromatin in late G1." Mol
Cell Biol 24(10): 4546-4556.
Tavazoie, S. F., C. Alarcon, et al. (2008). "Endogenous human microRNAs that suppress
breast cancer metastasis." Nature 451(7175): 147-152.
Taylor, W. R. and G. R. Stark (2001). "Regulation of the G2/M transition by p53."
Oncogene 20(15): 1803-1815.
Tazawa, H., N. Tsuchiya, et al. (2007). "Tumor-suppressive miR-34a induces
senescence-like growth arrest through modulation of the E2F pathway in human
colon cancer cells." Proc Natl Acad Sci U S A 104(39): 15472-15477.
te Poele, R. H., A. L. Okorokov, et al. (2002). "DNA damage is able to induce
senescence in tumor cells in vitro and in vivo." Cancer Res 62(6): 1876-1883.
Teh, M. T., S. T. Wong, et al. (2002). "FOXM1 is a downstream target of Gli1 in basal
cell carcinomas." Cancer Res 62(16): 4773-4780.
Tergaonkar, V. (2006). "NFkappaB pathway: a good signaling paradigm and therapeutic
target." Int J Biochem Cell Biol 38(10): 1647-1653.
Thomas, D. M., S. A. Carty, et al. (2001). "The retinoblastoma protein acts as a
transcriptional coactivator required for osteogenic differentiation." Mol Cell
8(2): 303-316.
Thomas, M., D. Pim, et al. (1999). "The role of the E6-p53 interaction in the molecular
pathogenesis of HPV." Oncogene 18(53): 7690-7700.
Tjian, R. and A. Robbins (1979). "Enzymatic activities associated with a purified simian
virus 40 T antigen-related protein." Proc Natl Acad Sci U S A 76(2): 610-614.
Tobaben, S., P. Thakur, et al. (2001). "A trimeric protein complex functions as a
synaptic chaperone machine." Neuron 31(6): 987-999.
Tolbert, D., X. Lu, et al. (2002). "p19(ARF) is dispensable for oncogenic stress-induced
p53-mediated apoptosis and tumor suppression in vivo." Mol Cell Biol 22(1):
370-377.
Trentin, J. J., Y. Yabe, et al. (1962). "The quest for human cancer viruses." Science 137:
835-841.
Trimarchi, J. M., B. Fairchild, et al. (1998). "E2F-6, a member of the E2F family that
can behave as a transcriptional repressor." Proc Natl Acad Sci U S A 95(6):
2850-2855.
Trimarchi, J. M., B. Fairchild, et al. (2001). "The E2F6 transcription factor is a
component of the mammalian Bmi1-containing polycomb complex." Proc Natl
Acad Sci U S A 98(4): 1519-1524.
Tsukamoto, K., N. Ito, et al. (1998). "Allelic loss on chromosome 1p is associated with
progression and lymph node metastasis of primary breast carcinoma." Cancer
82(2): 317-322.
Tyner, S. D., S. Venkatachalam, et al. (2002). "p53 mutant mice that display early
ageing-associated phenotypes." Nature 415(6867): 45-53.
van 't Veer, L. J., H. Dai, et al. (2002). "Gene expression profiling predicts clinical
outcome of breast cancer." Nature 415(6871): 530-536.
362
van der Lugt, N. M., J. Domen, et al. (1994). "Posterior transformation, neurological
abnormalities, and severe hematopoietic defects in mice with a targeted deletion
of the bmi-1 proto-oncogene." Genes Dev 8(7): 757-769.
Vasudevan, S. A., H. V. Russell, et al. (2007). "Neuroblastoma-derived secretory protein
messenger RNA levels correlate with high-risk neuroblastoma." J Pediatr Surg
42(1): 148-152.
Vaziri, N. D., Z. Ni, et al. (1998). "Upregulation of renal and vascular nitric oxide
synthase in young spontaneously hypertensive rats." Hypertension 31(6): 12481254.
Veldman, T., I. Horikawa, et al. (2001). "Transcriptional activation of the telomerase
hTERT gene by human papillomavirus type 16 E6 oncoprotein." J Virol 75(9):
4467-4472.
Vijg, J. (2000). "The science of aging and the need for a mechanistic approach." Mech
Ageing Dev 114(1): 1-3.
Vijg, J., R. A. Busuttil, et al. (2005). "Aging and genome maintenance." Ann N Y Acad
Sci 1055: 35-47.
Vogelstein, B., D. Lane, et al. (2000). "Surfing the p53 network." Nature 408(6810):
307-310.
Volinia, S., G. A. Calin, et al. (2006). "A microRNA expression signature of human
solid tumors defines cancer gene targets." Proc Natl Acad Sci U S A 103(7):
2257-2261.
von Zglinicki, T., G. Saretzki, et al. (2005). "Human cell senescence as a DNA damage
response." Mech Ageing Dev 126(1): 111-117.
Vooijs, M. and A. Berns (1999). "Developmental defects and tumor predisposition in Rb
mutant mice." Oncogene 18(38): 5293-5303.
Voorhoeve, P. M., C. le Sage, et al. (2006). "A genetic screen implicates miRNA-372
and miRNA-373 as oncogenes in testicular germ cell tumors." Cell 124(6): 11691181.
Vousden, K. H. and X. Lu (2002). "Live or let die: the cell's response to p53." Nat Rev
Cancer 2(8): 594-604.
Vousden, K. H., B. Vojtesek, et al. (1993). "HPV-16 E7 or adenovirus E1A can
overcome the growth arrest of cells immortalized with a temperature-sensitive
p53." Oncogene 8(6): 1697-1702.
Wagner, M., B. Hampel, et al. (2001). "Replicative senescence of human endothelial
cells in vitro involves G1 arrest, polyploidization and senescence-associated
apoptosis." Exp Gerontol 36(8): 1327-1347.
Wagner, W., P. Horn, et al. (2008). "Replicative senescence of mesenchymal stem cells:
a continuous and organized process." PLoS One 3(5): e2213.
Wajapeyee, N., R. W. Serra, et al. (2008). "Oncogenic BRAF induces senescence and
apoptosis through pathways mediated by the secreted protein IGFBP7." Cell
132(3): 363-374.
Walia, V., M. Ding, et al. (2009). "hCLCA2 Is a p53-Inducible Inhibitor of Breast
Cancer Cell Proliferation." Cancer Res 69(16): 6624-6632.
Wang, H., Q. Zhang, et al. (2003). "hSGT interacts with the N-terminal region of
myostatin." Biochem Biophys Res Commun 311(4): 877-883.
363
Wang, H. G., E. Moran, et al. (1995). "E1A promotes association between p300 and
pRB in multimeric complexes required for normal biological activity." J Virol
69(12): 7917-7924.
Wang, I. C., Y. J. Chen, et al. (2005). "Forkhead box M1 regulates the transcriptional
network of genes essential for mitotic progression and genes encoding the SCF
(Skp2-Cks1) ubiquitin ligase." Mol Cell Biol 25(24): 10875-10894.
Wang, J., N. K. Jacob, et al. (2009). "RelA/p65 functions to maintain cellular senescence
by regulating genomic stability and DNA repair." EMBO Rep 10(11): 12721278.
Wang, R. H., K. Sengupta, et al. (2008). "Impaired DNA damage response, genome
instability, and tumorigenesis in SIRT1 mutant mice." Cancer Cell 14(4): 312323.
Wang, W., J. X. Chen, et al. (2002). "Sequential activation of the MEK-extracellular
signal-regulated kinase and MKK3/6-p38 mitogen-activated protein kinase
pathways mediates oncogenic ras-induced premature senescence." Mol Cell Biol
22(10): 3389-3403.
Wang, Y. and M. You (2001). "Alternative splicing of the K-ras gene in mouse tissues
and cell lines." Exp Lung Res 27(3): 255-267.
Waragai, M., S. Nagamitsu, et al. (2006). "Ataxin 10 induces neuritogenesis via
interaction with G-protein beta2 subunit." J Neurosci Res 83(7): 1170-1178.
Watts, L. M., V. P. Manchem, et al. (2005). "Reduction of hepatic and adipose tissue
glucocorticoid receptor expression with antisense oligonucleotides improves
hyperglycemia and hyperlipidemia in diabetic rodents without causing systemic
glucocorticoid antagonism." Diabetes 54(6): 1846-1853.
Weber, J. D., J. R. Jeffers, et al. (2000). "p53-independent functions of the p19(ARF)
tumor suppressor." Genes Dev 14(18): 2358-2365.
Weber, J. D., L. J. Taylor, et al. (1999). "Nucleolar Arf sequesters Mdm2 and activates
p53." Nat Cell Biol 1(1): 20-26.
Webley, K., J. A. Bond, et al. (2000). "Posttranslational modifications of p53 in
replicative senescence overlapping but distinct from those induced by DNA
damage." Mol Cell Biol 20(8): 2803-2808.
Wei, S. and J. M. Sedivy (1999). "Expression of catalytically active telomerase does not
prevent premature senescence caused by overexpression of oncogenic Ha-Ras in
normal human fibroblasts." Cancer Res 59(7): 1539-1543.
Wei, W., R. M. Hemmer, et al. (2001). "Role of p14(ARF) in replicative and induced
senescence of human fibroblasts." Mol Cell Biol 21(20): 6748-6757.
Wei, W., U. Herbig, et al. (2003). "Loss of retinoblastoma but not p16 function allows
bypass of replicative senescence in human fibroblasts." EMBO Rep 4(11): 10611066.
Weinberg, R. A. (1995). "The retinoblastoma protein and cell cycle control." Cell 81(3):
323-330.
Weinberg, W. C. and S. H. Yuspa (1997). "An antibody to p53 recognizes soluble
keratins in epidermal keratinocyte cultures under differentiating, but not
proliferating, conditions." J Invest Dermatol 109(4): 611-612.
Weinrich, S. L., R. Pruzan, et al. (1997). "Reconstitution of human telomerase with the
template RNA component hTR and the catalytic protein subunit hTRT." Nat
Genet 17(4): 498-502.
364
Wellinger, R. J. and D. Sen (1997). "The DNA structures at the ends of eukaryotic
chromosomes." Eur J Cancer 33(5): 735-749.
West, M. D., O. M. Pereira-Smith, et al. (1989). "Replicative senescence of human skin
fibroblasts correlates with a loss of regulation and overexpression of collagenase
activity." Exp Cell Res 184(1): 138-147.
Westbrook, T. F., F. Stegmeier, et al. (2005). "Dissecting cancer pathways and
vulnerabilities with RNAi." Cold Spring Harb Symp Quant Biol 70: 435-444.
White, E., S. K. Chiou, et al. (1994). "Control of p53-dependent apoptosis by E1B, Bcl2, and Ha-ras proteins." Cold Spring Harb Symp Quant Biol 59: 395-402.
White, E., P. Sabbatini, et al. (1992). "The 19-kilodalton adenovirus E1B transforming
protein inhibits programmed cell death and prevents cytolysis by tumor necrosis
factor alpha." Mol Cell Biol 12(6): 2570-2580.
Whitfield, M. L., G. Sherlock, et al. (2002). "Identification of genes periodically
expressed in the human cell cycle and their expression in tumors." Mol Biol Cell
13(6): 1977-2000.
Whyte, P., K. J. Buchkovich, et al. (1988). "Association between an oncogene and an
anti-oncogene: the adenovirus E1A proteins bind to the retinoblastoma gene
product." Nature 334(6178): 124-129.
Whyte, P., N. M. Williamson, et al. (1989). "Cellular targets for transformation by the
adenovirus E1A proteins." Cell 56(1): 67-75.
Wierstra, I. and J. Alves (2007). "FOXM1, a typical proliferation-associated
transcription factor." Biol Chem 388(12): 1257-1274.
Wilfred, B. R., W. X. Wang, et al. (2007). "Energizing miRNA research: a review of the
role of miRNAs in lipid metabolism, with a prediction that miR-103/107
regulates human metabolic pathways." Mol Genet Metab 91(3): 209-217.
Williams, G. L., T. M. Roberts, et al. (2007). "Bub1: escapades in a cellular world." Cell
Cycle 6(14): 1699-1704.
Wolyniec, K., S. Wotton, et al. (2009). "RUNX1 and its fusion oncoprotein derivative,
RUNX1-ETO, induce senescence-like growth arrest independently of replicative
stress." Oncogene 28(27): 2502-2512.
Woods, D. B. and K. H. Vousden (2001). "Regulation of p53 function." Exp Cell Res
264(1): 56-66.
Wotton, S. F., K. Blyth, et al. (2004). "RUNX1 transformation of primary embryonic
fibroblasts is revealed in the absence of p53." Oncogene 23(32): 5476-5486.
Wu, S. J., F. H. Liu, et al. (2001). "Different combinations of the heat-shock cognate
protein 70 (hsc70) C-terminal functional groups are utilized to interact with
distinct tetratricopeptide repeat-containing proteins." Biochem J 359(Pt 2): 419426.
Wu, W., Z. Lin, et al. (2009). "Expression profile of mammalian microRNAs in
endometrioid adenocarcinoma." Eur J Cancer Prev 18(1): 50-55.
Xiao, Z. X., J. Chen, et al. (1995). "Interaction between the retinoblastoma protein and
the oncoprotein MDM2." Nature 375(6533): 694-698.
Yamakuchi, M., M. Ferlito, et al. (2008). "miR-34a repression of SIRT1 regulates
apoptosis." Proc Natl Acad Sci U S A 105(36): 13421-13426.
Yan, D., X. Zhou, et al. (2009). "MicroRNA-34a inhibits uveal melanoma cell
proliferation and migration through downregulation of c-Met." Invest
Ophthalmol Vis Sci 50(4): 1559-1565.
365
Yang, A., M. Kaghad, et al. (2002). "On the shoulders of giants: p63, p73 and the rise of
p53." Trends Genet 18(2): 90-95.
Yang, A., M. Kaghad, et al. (1998). "p63, a p53 homolog at 3q27-29, encodes multiple
products with transactivating, death-inducing, and dominant-negative activities."
Mol Cell 2(3): 305-316.
Yao, J., L. Duan, et al. (2007). "Overexpression of BLCAP induces S phase arrest and
apoptosis independent of p53 and NF-kappaB in human tongue carcinoma :
BLCAP overexpression induces S phase arrest and apoptosis." Mol Cell
Biochem 297(1-2): 81-92.
Yeung, F., J. E. Hoberg, et al. (2004). "Modulation of NF-kappaB-dependent
transcription and cell survival by the SIRT1 deacetylase." EMBO J 23(12): 23692380.
Yew, P. R., X. Liu, et al. (1994). "Adenovirus E1B oncoprotein tethers a transcriptional
repression domain to p53." Genes Dev 8(2): 190-202.
Yoneda-Kato, N. and J. Y. Kato (2008). "Shuttling imbalance of MLF1 results in p53
instability and increases susceptibility to oncogenic transformation." Mol Cell
Biol 28(1): 422-434.
Yoneda-Kato, N., K. Tomoda, et al. (2005). "Myeloid leukemia factor 1 regulates p53
by suppressing COP1 via COP9 signalosome subunit 3." EMBO J 24(9): 17391749.
Yoon, I. K., H. K. Kim, et al. (2004). "Exploration of replicative senescence-associated
genes in human dermal fibroblasts by cDNA microarray technology." Exp
Gerontol 39(9): 1369-1378.
Young, A. P. and G. D. Longmore (2004). "Ras protects Rb family null fibroblasts from
cell death: a role for AP-1." J Biol Chem 279(12): 10931-10938.
Zellweger, T., H. Miyake, et al. (2001). "Chemosensitization of human renal cell cancer
using antisense oligonucleotides targeting the antiapoptotic gene clusterin."
Neoplasia 3(4): 360-367.
Zender, L., W. Xue, et al. (2008). "An oncogenomics-based in vivo RNAi screen
identifies tumor suppressors in liver cancer." Cell 135(5): 852-864.
Zhang, H. S., A. A. Postigo, et al. (1999). "Active transcriptional repression by the RbE2F complex mediates G1 arrest triggered by p16INK4a, TGFbeta, and contact
inhibition." Cell 97(1): 53-61.
Zhang, S., R. Binari, et al. (2010). "A genomewide RNA interference screen for
modifiers of aggregates formation by mutant Huntingtin in Drosophila." Genetics
184(4): 1165-1179.
Zhang, Y., M. Li, et al. (2009). "Profiling of 95 microRNAs in pancreatic cancer cell
lines and surgical specimens by real-time PCR analysis." World J Surg 33(4):
698-709.
Zhang, Y., Y. Xiong, et al. (1998). "ARF promotes MDM2 degradation and stabilizes
p53: ARF-INK4a locus deletion impairs both the Rb and p53 tumor suppression
pathways." Cell 92(6): 725-734.
Zhao, J. J., J. Yang, et al. (2009). "Identification of miRNAs associated with
tumorigenesis of retinoblastoma by miRNA microarray analysis." Childs Nerv
Syst 25(1): 13-20.
366
Zheng, W., H. Wang, et al. (2004). "Regulation of cellular senescence and p16(INK4a)
expression by Id1 and E47 proteins in human diploid fibroblast." J Biol Chem
279(30): 31524-31532.
Zhou, L., X. Qi, et al. (2008). "MicroRNAs miR-186 and miR-150 down-regulate
expression of the pro-apoptotic purinergic P2X7 receptor by activation of
instability sites at the 3'-untranslated region of the gene that decrease steady-state
levels of the transcript." J Biol Chem 283(42): 28274-28286.
Zhu, J., D. Woods, et al. (1998). "Senescence of human fibroblasts induced by
oncogenic Raf." Genes Dev 12(19): 2997-3007.
Zhu, J. W., D. DeRyckere, et al. (1999). "A role for E2F1 in the induction of ARF, p53,
and apoptosis during thymic negative selection." Cell Growth Differ 10(12):
829-838.
Zhu, J. Y., P. W. Rice, et al. (1991). "Mapping the transcriptional transactivation
function of simian virus 40 large T antigen." J Virol 65(6): 2778-2790.
Zhu, L., S. van den Heuvel, et al. (1993). "Inhibition of cell proliferation by p107, a
relative of the retinoblastoma protein." Genes Dev 7(7A): 1111-1125.
Zimmers, T. A., M. V. Davies, et al. (2002). "Induction of cachexia in mice by
systemically administered myostatin." Science 296(5572): 1486-1488.
Zindy, F., C. M. Eischen, et al. (1998). "Myc signaling via the ARF tumor suppressor
regulates p53-dependent apoptosis and immortalization." Genes Dev 12(15):
2424-2433.
Zindy, F., H. Soares, et al. (1997). "Expression of INK4 inhibitors of cyclin Ddependent kinases during mouse brain development." Cell Growth Differ 8(11):
1139-1150.
Zuo, L., J. Weger, et al. (1996). "Germline mutations in the p16INK4a binding domain
of CDK4 in familial melanoma." Nat Genet 12(1): 97-99.
Zuo, Z., M. Zhao, et al. (2006). "Functional analysis of bladder cancer-related protein
gene: a putative cervical cancer tumor suppressor gene in cervical carcinoma."
Tumour Biol 27(4): 221-226.
zur Hausen, H. (2001). "Cervical carcinoma and human papillomavirus: on the road to
preventing a major human cancer." J Natl Cancer Inst 93(4): 252-253.
367